I met Eric Partaker yesterday. Former Skype exec, runs a 200k+ subscriber newsletter. I was surprised when the topic turned to him asking about using Skyp for his webinar follow-ups. He runs about 20 webinars a quarter, gets hundreds of registrations per webinar, and wants to do one-to-one follow-up at scale. Classic use case for what we built–with maybe more people than most people get on webinars.

During this discussion, he asked a simple question of me more broadly: "What are you optimizing for?"

I started listing things. Skyp's growth. The advisory practice. The newsletter. Strategic partnerships. Each answer was technically correct. Each was also a dodge.

The conversation forced me to confront something I'd been avoiding. I wasn't building a focused company. I was maintaining a portfolio of related but distinct projects, each claiming to support the others. The advisory work informs the product. The personal brand drives awareness. The various content channels reach different audiences.

I was optimizing for optionality, not outcomes. Which meant I was not optimizing at all.

The Tyranny of Options

Founders love optionality. Keep the advisory work to pay the bills, and in case Skyp doesn't scale. Maintain the personal brand in case the company pivots. Write multiple newsletters to hedge our bets across audiences. Build relationships in three different directions in case one doesn't work out.

This isn't stupid. Markets change. Products fail (I will never forget). Hedging makes rational sense when you're navigating uncertainty.

One huge plus of raising venture capital is that it enables you–if not compels you–to put things down. With a salary, there’s no need for a side hustle. With a board, there’s no hiding from missed KPIs.

Optionality has a hidden cost that compounds daily. It’s the cognitive load you carry constantly—the mental overhead of maintaining multiple contexts, multiple identities, multiple stories about what you're building.

I had known I’d been paying this price for months. Every time I do a newsletter collab I write a byline–and it’s hard to keep it short. Because it isn’t. Founder/CEO of Skyp. Also advisor. Also newsletter author. What do I drive people to? Skyp? This newsletter? That newsletter?

Nothing was growing well because my own marketing wasn’t clear. I was treating Skyp like one priority among several. Writing content about general startup topics–that mentioned an email outreach product. Meeting potential advisory clients that might, actually, be good Skyp customers–but not bringing that up, because it was out of context. Building my LinkedIn presence as a general startup expert rather than specifically around GTM as the founder of an outbound tool. All reasonable activities. Just not focused.

The moment I admitted this to myself, something unexpected happened. I felt lighter.

Choosing what not to do isn't constraint. It's clarifying.

Focus as Freedom

Conventional wisdom says focus means constraint. You close doors. You narrow options. You limit yourself.

But I've realized the opposite is true. Focus creates freedom.

When you decide what you're actually building, you stop spending energy on maintaining facades. You stop splitting attention across multiple identities. You stop hedging bets that don't need hedging.

All that energy goes somewhere useful instead.

This is even true within a startup. When you put down that idea and allow yourself to just work on the core product, one or two features instead of all of them, you just feel better. Lighter, more focused, less crushed by the work.

So here's what I'm doing: I'm going all in on Skyp.

The advisory work continues with existing clients. It informs the product and I genuinely enjoy the work. But it's not the main thing anymore, and I’m not looking to grow it. Watch, I’ll get 3 referrals from this email.

The LinkedIn presence becomes a focused Skyp channel. Everything I write will drive awareness of the problem we solve, demonstrate expertise in the space we operate, or share lessons from building the company. If it doesn't serve one of those purposes, I shouldn't be writing it.

And I’ll give myself the freedom to invest the time so everything I write will be be a banger. By not having to do other stuff. Marketing is the work, sometimes, and this is the marketing.

The newsletters consolidate. Instead of maintaining separate identities—the thoughtful GTM advisor writing Seed to Sequoia, the company founder writing Skyp content—I am integrating them. This will help me clarify the messaging for those collab by-lines: Find Alex at Skyp. Period.

What This Means for You

If you're reading this, you're on the Seed to Sequoia newsletter. I've been writing it for founders, operators and VCs navigating early-stage growth for almost two years now. It's been personal, occasionally random, always authentic. I'm keeping it.

But I'm also moving everyone to the Skyp newsletter.

The Skyp newsletter is more tactical. It focuses specifically on outbound, sales enablement, and GTM execution—the things our product actually helps with. If you're building a sales motion, testing outbound channels, or trying to figure out how to generate pipeline without hiring an SDR team, you'll find it useful.

You might also find you don't care about those topics. That's fine. Unsubscribe. I mean that genuinely. Attention is the most valuable thing you have. Don't waste it on content that doesn't serve you.

The Skyp newsletter is designed to give people what they really want–not what I want to write, though obviously that has some influence. Really useful stuff. Which is why it gets a higher open rate than this newsletter–which already gets a very high open rate–and I get as many responses to each edition as I do here, with far fewer readers. In other words, it hits. I think you’ll like it.

Seed to Sequoia will continue, but less frequently. When I have something to say that doesn't fit the Skyp frame—observations about business, partnerships, the occasionally spicy take on market dynamics—I'll write it here. Think of this as the authentically random newsletter for insiders. The Skyp newsletter is the focused one, for everyone.

The Ask

Skyp is at around $10k MRR. Our goal is to more than double that by year-end.

When we hit $10k at the end of August, I was shocked. We set a goal to hit $25k by Christmas, which felt very doable. We had momentum. The product was working. Customers were seeing results. Doubling felt like a matter of execution, not luck.

Now it's mid-November. The goal is still within reach, but it's no longer comfortable. We need roughly $12-15k in new MRR to hit the target, which means closing 5-12 new customers in the next six weeks. Possible, but not guaranteed, especially with holidays coming up. The difference between hitting it and missing it is probably five conversations that go well.

Which means I need your help.

If you know someone building an outbound motion—founders doing their own sales, ops leaders trying to scale without hiring, revenue teams that need their reps to be more productive—share Skyp with them. Send them to skyp.ai. Make an intro. Forward this email. Whatever fits your style.

If you run outbound yourself and haven't tried the product, book a demo at skyp.ai/demo. We'll get you set up in 15 minutes. First campaign can be live same day. I did 4 demos last week; 2 people signed up and one is going through procurement. I like that conversion rate, just need to do more demos!

The product works. People who use it see results. We're not struggling because of the product. We're struggling because not enough people who have the problem know we exist.

That's solvable. But it requires more than me showing up on LinkedIn. It requires people telling others about it.

So if you've gotten value from this newsletter over the years—or even just this one post—here's how you can return it: tell one person about Skyp who has the problem we solve.

One person. That's the ask.

Focus Yourself

This time of year invites reflection. The year is ending. New planning cycles are starting. Companies do retrospectives. Investors review portfolios. Founders assess what's working.

You don't need a formal planning process to ask useful questions. In fact, formal planning processes sometimes obscure the real answers. They create structure that feels productive while avoiding the uncomfortable truths you actually need to confront. They put company before self.

So ask yourself: What's serving you? What isn't?

Not "what's interesting" or "what might be valuable someday." What is actively serving your most important goal right now?

What are you maintaining out of obligation rather than intention? What projects are you keeping alive because you spent time on them, not because they're the best use of your time going forward? What relationships are you preserving that don't generate anything meaningful? What content are you creating that nobody reads?

The sunk cost fallacy applies to more than investments. It applies to activities, identities, and commitments. Just because you've been doing something for a long time doesn't mean you should keep doing it.

What would you stop doing if you were honest about what actually matters?

This is harder than it sounds. We construct elaborate justifications for why everything we do is essential. We tell stories about how unrelated activities all support each other. We convince ourselves that maintaining optionality is strategic.

I am doing it right now.

Sometimes it’s true. Usually it's not.

Give yourself permission to stop things. To narrow. To commit. Not because focus is virtuous in some abstract sense, but because it's effective in a very practical one.

Every successful company I've studied had a moment where the founders went all in. They stopped hedging. They stopped maintaining backup plans. They committed completely to one direction and accepted that if it didn't work, they'd deal with that when it happened.

That's uncomfortable. But discomfort isn't a signal you're doing something wrong. Often it's a signal you're doing something that matters.

The insight that creates freedom isn't that you should do more. It's that you can do less.

And that less might be exactly what you need to break through.

Want the tactical version of this newsletter? Join the Skyp newsletter for weekly posts on outbound, sales, and GTM.

Ready to scale your outbound without hiring an SDR team? Book a demo.

ASK: I’m looking to talk to people in private equity or at PE backed companies about GTM. If that’s you, just reply! Or find a time. Not selling anything, just seeking knowledge. And sharing knowledge, if desired.

OpenAI’s monster forecast

OpenAI projects $200 billion in revenue by 2030, burning $115 billion to get there. Presumably this is in preparation to raise money–perhaps even IPO. Here’s a chart, in case picturing 2024’s $3.7 billion in revenue growing over 5,000% in 5 years is difficult:

OpenAI forecast, courtesy of Titan

To put that in perspective, that is almost as much as Microsoft’s revenue today ($282 billion in 2025). I get that growth happens faster today, but it took Microsoft 43 years to get there. OpenAI is going to do it in roughly five. How?

Where is the revenue coming from?

According to internal documents obtained by The Wall Street Journal and The Information, OpenAI has a plan. By 2030, they expect:

• ~$90 billion from ChatGPT subscriptions (up from $10 billion ARR in 2025)

• ~$29 billion from AI agents—products like "Operator" priced at $2,000 to $20,000 per month

• ~$25-110 billion from monetizing free users through ads and affiliate commerce

• ~$22 billion from API/enterprise sales (up from $2 billion)

Let's start with the idea that companies will suddenly spend massive amounts on AI agents and enterprise software. From 2029 to 2030, OpenAI grows by around $45 billion—which is more than all of Salesforce's 2025 revenue. Is the idea that people just turn off Salesforce, and move all that spending over to OpenAI?

Or maybe they shut off Oracle, which did $57 billion in revenue in 2025?

If they shut off both, voila–$95 billion right there. Given the stickiness of CRM and ERP software I find that exceptionally unlikely. So the revenue would have to come from somewhere else–and possibly eclipse what companies spend on their systems of record.

Typically large companies spend 2-3% of revenue on ERP or CRM. Were OpenAI to replace that spend, or get an incremental 2-3% spend, it would imply that OpenAI is serving companies doing a combined ~$2 trillion in revenue–representing, pick your relationship, 10% of the Fortune 500 by revenue or 7% of US GDP (based on $51 billion in OpenAI enterprise revenue).

I’m dubious. But Open AI says only about a quarter of the revenue will come from enterprise–$29 billion from AI agents and $22 from API and enterprise sales. Where does the other half come from? Consumers? Let's dig in there.

What about consumer?

If a quarter of OpenAI’s revenue comes from enterprise, that leaves $150 billion to come from consumers.

OpenAI says $90 billion will come from ChatGPT subscriptions. But the rate of paying there is only 5%. So while OpenAI forecasts 9x growth–up from $10 billion today–how exactly would that work? At some point you run out of people who can afford a $20 or $200 monthly subscription.

There are companies who sell subscriptions en masse. $90 billion in subscription revenue would be more revenue than Amazon Prime and Netflix globally–combined. Sure, OpenAI has five years to get there. It only took Amazon 20 and Netflix 27 years. Cake walk?

Let’s make it easy so we can move on: every Amazon Prime subscriber quits and spends that money on ChatGPT. That’s $44 billion. Due to inflation, we round up to $50.

Now we just have to find the other $100 billion somehow.

If OpenAI consumes all the consumer markets–how much revenue is there? I don’t mean affiliate links, I mean just take all the revenue. Let’s look at the two biggest that it’s most likely to go after: travel and ecommerce.

In travel, Booking Holdings does $24 billion in revenue, Expedia $14 billion, and Airbnb $11 billion. So if OpenAI just absorbed all of that revenue, and those companies (or the consumers who have been using them for years) just stopped–that’s $49 billion. Those companies go away. Consumers only buy travel through OpenAI. Nobody puts up a fight, or retains any share.

I don’t buy all of it–maybe some? Half? That’d be $25 billion, so that leaves $75 billion from ecommerce.

Headline: AI takes over ecommerce. This one is tricky, as it’s more than pure software. There are pesky warehouses, logistics hubs, and trucks with “Amazon” written on them to be worried about.

Let’s simplify and start with software. Shopify–one of Canada’s biggest tech successes ever–did around $9 billion in 2024 revenue, roughly 3% of GMV on the platform (around $300 billion). If ChatGPT can convince consumers to buy from it instead of Shopify, including hosting their websites and handling all of the backend of running a retail business, let’s say it gets that full $9 billion.

We’re down to just $66 billion to find.

How could I forget about eBay and Rakuten? $10 billion in annual revenue each, roughly. We’ll absorb that, also. Let’s not ask questions about servicing their revenue–just take it, like 100% of it is affiliate fees paid to OpenAI.

$46 billion left. There are others (Coupang, Pinduodou) but they’re more international so we’ll leave them alone for now. There’s also Etsy but at ~$2 bn it’s a rounding error. What about Amazon?

In the US Amazon has the lion’s share of ecommerce revenue–$450 billion. There’s massive capital cost–warehouses, logistics, and labor–and low margins. Not the kind of margins in OpenAi’s forecast for 2030. Also, scale. There’s no way to take just an eighth of Amazon’s revenue–there’d be no same day delivery!

Instead, could OpenAI get Amazon to pay it a referral fee?

No. No it will not. Amazon is not run by fools.

And even if it did–would it be to the tune of our missing $46 billion–a tenth of Amazon’s ecommerce revenue? Ha!

Putting this all together, if OpenAI eats the entire publicly traded travel market and all of Shopify’s revenue, but nothing from Amazon, it still comes up short–$46 billion short (not to backtrack, but that’s more than Salesforce’s revenue and 10x Shopify’s).

Could ads close the gap? Maybe. Thing is if Reddit has long struggled to figure out advertising, partially due to how unpredictable the content is, I’m not sure ChatGPT–much of which is based on Reddit content, with added hallucinations–is going to appeal to advertisers on such a tight timeframe. Unlike consumers, advertisers don’t spend billions on a whim. It took Google 4 years to hit $1 billion, but those were different times.

But hey, challenging but maybe doable? What could possibly go wrong?

The other elephant

Revenue growth does not happen in a vacuum. While OpenAI is planning to consume roughly 10% of the software market, it is not without competitiors. Anthropic is a not-so-distant number two.

Anthropic forecast, courtesy of Titan

Anthropic does not compete as aggressively in consumer. Where OpenAI has gone after the consumer market with a vengeance, Anthropic has focused more on enterprise. It is not nearly the same household name, even though it’s better at certain tasks consumers use AI for–at least in my experience. It would be foolish to assume that Anthropic doesn’t eventually take a bite of the consumer market, but that’s not the main threat it poses to the forecast.

Enterprise is where Anthropic poses the biggest risk to OpenAI’s forecast. Anthropic generates 80% of its revenue from enterprise customers–and has more marketshare in enterprise than OpenAI. So when the question is, will OpenAI get all of the enterprise agent budget, and all of the market for AI APIs? The answer is “no” with nearly 100% confidence. Because Anthropic is already beating it there.

Anthropic is also doing so in anticipation of profitability much sooner, but why would we talk about that?

The other other elephant

Did I forget to mention Gemini? They only have 20% of the enterprise market anyway, and just ~90% of the consumer search market.

I’m sure Sam knows something I don’t: specifically, that Google will just roll over and let OpenAI take share in search, ads and enterprise business. And while OpenAI spends $115 billion to achieve its full potential over the next five years, Google won’t do anything to fight back with the ~$72 billion it generates in free cash flow every year.

Smaller question: To profit or not to profit?

The revenue goals seem a bit of a stretch. What about the cost side of this equation? Is $200 billion in phantom revenue worth $115 billion in burn to achieve?

Veteran entrepreneurs point to how AI has transformed starting a company. You can do so much more with so many fewer people–and less capital, in general–that it feels like the startup old days. When raising $1mm was a seed round, not a joke. When you could feed your whole team with pizza–for most meals, most days.

And that’s what’s so weird about this forecast. Back in the old days, startups were profitable almost from the get go. Microsoft was profitable from the beginning. Their motto could be “Printing Cash Since 1982”. Google was profitable in year 3, and throwing off hundreds of millions by the time it went public in 2004.

While you might argue that OpenAI is different–it isn’t, actually. Google was a famously server-intense business. But instead of spending ahead of demand, it built its own servers from reject parts it bought on the cheap from manufacturers. It had a team of engineers building BIOS (the software that makes all of the parts work, usually provided by the manufacturer) because the manufacturer had abandoned them. It was a lot cheaper to buy the scrap parts for pennies on the dollar and pay very expensive developers to build software to run them than it was to pay for top of the line hardware.

OpenAI seems happy to just pony up for the top of the line hardware. And datacenters to hold them. And power to run them. With seeming disregard for costs. While spending on even more expensive, less profitable undertakings like image and voice applications. Even though it could easily tweak the settings in ChatGPT to burn less compute, and be slightly closer to profitable. But what is, say, 5% of $115 billion, anyway? Other than roughly the market cap of Etsy.

Bigger than Microsoft

The forecast puts OpenAI close to Microsoft’s revenue today–$281 billion in 2025–just five years out. That means forecasting an increase of 5,400% from $3.7 billion in revenue in 2024 to ~$200 billion in 2030. On 2 years of data. Has any company ever done that? I’m sure many have forecast 50x growth on meager data–I mean actually achieved it.

Microsoft has had an impressive run of growth. The key difference is Microsoft was always profitable. In 2025 it generated $102 billion in profit. (Mere coincidence that that is in the ballpark of how much OpenAI plans on torching to achieve its revenue forecast?)

The revolution AI represents seems at least on the same order of magnitude as the revolution Apple and Microsoft ushered in with the rise of personal computing. Maybe I’m old, and a curmudgeon, but I don’t see how this revolution happens 5x faster or is that much bigger. Or, why one company needs to spend so much money to get there.

Why?

The risks of overcapitalizing and overspending are real, whether you’re a small company raising your first venture round or operating at the scale of OpenAI. The risk is always most acute in that early phase of the hype cycle.

Remember Elon and his “full self driving” talk? That started in 2013. Yes, over a decade ago. I wanted to believe him. Lots of people wanted to believe him. And did believe him. Since then, we’ve had one major lawsuit about it and recently got a handful of Tesla robotaxis in San Francisco (of course all rides cost $4.20–yes, that’s a fact).

The question I ask is: why? Why compels OpenAI to grow this quickly? Why do they have to spend all of this money? What is so urgent here?

Some people have been using AI for the last ~12 months. Most people still aren’t really using it anywhere near its potential. Do we really need to scale this quickly? Is this the reason–that everyone else will start to use it?

And what if all of that demand doesn’t materialize, then what? Will OpenAI turn into the next WebVan, overspending on distribution centers until it went bankrupt?

Meanwhile startups building on AI tools–like Skyp–are more efficient and profitable than ever. I realize there might be bigger problems to solve than writing better sales emails (maybe?) but there’s absolutely no reason to subsidize usage to the tune of a hundred billion dollars to get there. Usage that has spiked–but may unspike, one never knows.

The real question

We've been talking about—and waiting for—AGI for 50 years. What's the rush now?

OpenAI is betting $115 billion that demand will materialize at a pace we've never seen before. That enterprise software buyers will abandon sticky systems of record. That consumers will shift all their shopping and travel through ChatGPT. That this happens fast enough to justify the spend.

Maybe the real bet is something bigger: that OpenAI creates an entirely new category where companies that were spending 2-3% of revenue on core systems now spend 4-6%. Which seems almost reasonable—until you realize that companies currently spend roughly 25% of revenue on human labor. If AI is truly replacing humans at scale, OpenAI would be capturing a fraction of that spend.

And that raises a different question: what happens to the humans? And if AI isn't replacing humans at that scale, where exactly does $200 billion in new software spend come from?

Meanwhile, startups building on AI—like Skyp—are more capital efficient than software companies have been in decades. You can do more with less. Which makes the scale of OpenAI's bet even stranger.

The irony is that AI's biggest promise is efficiency. Yet OpenAI is pursuing the most capital-intensive path possible to get there—spending more to reach Microsoft's revenue in 5 years than Microsoft spent to build it over 43.

History suggests two possibilities: either this forecast is Elon-grade puffery designed to justify massive capital raises, or we're watching a potential WebVan or British Satellite Broadcasting moment for a technology with genuine transformative potential.

Either way, if Sam's selling shares at a valuation tied to this forecast, they're richly priced.

Thanks for reading! If you liked this please let me know–just reply. I’m figuring out what exactly to write about, and what you find most useful. This was fun to write–but was it useful? Or would more how-to content be better? You writing in is the only way I can know what you want to read about! I read every. Single. Email.

I've worked with dozens of stuck startups over the past few years. $1mm in ARR, $3mm, $15mm—but growth has stalled. Unfortunately, this happens to almost everybody. They all ask the same question: what do you do when growth stalls?

This is meant to be the definitive piece on finding the bugs in your GTM. Hopefully it's useful to you—whatever stage you're at.

When to Debug Your GTM

You can debug your GTM anytime. But there are times when it's necessary and the best use of your time as a founder:

• When growth slows. If your growth rate drops significantly quarter-over-quarter (say, 20% or more), it's time to investigate.

• When your top of funnel changes for the worse. A sudden drop in lead volume or quality demands immediate attention.

• When your close rate changes. If fewer opportunities are converting to customers, something in your process has broken.

• When you have high turnover in sales. Salespeople leave when they can't hit quota—often a sign of deeper GTM issues.

Where to start

I always follow the path of the customer's journey when trying to figure out what's wrong with GTM. This means starting at the awareness stage—where they become aware of a pain that your company solves—and moving through the marketing and sales process to closing and eventually retention.

Obviously this involves some assessment of product market fit (PMF). PMF is not a permanent state. You can have PMF and then things change. Maybe the change was something you did, like Blackberry rejecting touch screens and focusing on the enterprise while Google and Apple did the opposite. Or maybe it's something external—like Docker, which lost a key piece of the value chain to Google's Kubernetes and struggled to monetize ever since.

We'll get into PMF, but there's a fundamental assumption that you have or had PMF throughout, and the issue is more in your GTM than in your product. That's a big assumption, so beware. If you're seeing problems across multiple parts of the funnel, the issue might be PMF itself, not GTM execution. There are a lot more variables to consider if you’ve never really figured out GTM–I’ll touch on them throughout.

Top of Funnel GTM

A good reason to start at the beginning of the customer journey is that's where most GTM problems originate. Think about your future customer at this stage:

• They may or may not know they have a problem

• They may or may not know that a solution to the problem exists

• They may or may not have any budget dedicated to solving this problem

Some markets overcome this issue with just the sheer magical nature of their solutions. Building entire websites with a one-sentence instruction does that—which is why AI coding tools have seen such rapid adoption.

Others have struggled. Blockchain and crypto were hot in 2021-22, but faded fast (they’re back, kind of). Virtual reality has consistently failed to deliver on the hype. Meal kit delivery was on fire in 2015-2018–Blue Apron IPO’d–but as the market became saturated, customer acquisition costs rose and share prices fell (90%, in Blue Apron’s case). I could go on.

Talk to prospective customers—or, if you're talking to them already, go back and mine the call recordings. Is what you sell, or the problems you solve, a top priority for them? Do they have budget for it? Just because you think it should be a priority doesn't make it so.

Problems in the Marketing Top of Funnel

Once you've figured out whether your problem still has a market—because it might not, times change—then dig into the actual details. A good way to think of them is by channel. Not everything has to be perfect, but was something working that isn't working anymore?

Owned Channels

Your website. Are you getting the same traffic? Is it the same quality?

Social media. Are you getting the same engagement? Did algorithms change? Did audience preferences shift? If this is a primary channel, know that platforms evolve quickly.

Paid Channels

Paid marketing. Any change to performance? New entrants can drive up costs while reducing conversions. Check your cost-per-acquisition and conversion rates over time.

Earned and Relationship Channels

Outbound. Was your outbound working previously but not anymore? Did something change specifically with you (deliverability, messaging) or with the market? Google and Microsoft are constantly changing the deliverability rules–helps to have someone in your corner.

Lead nurturing gaps. Are you dropping leads on the floor? Someone interacts with your LinkedIn post—then what? Do they get enriched and engaged, or do they fall into a black hole?

Third-party references. Do you have mentions in articles, other people's social posts, etc., and has anything there changed?

Events. Did your event attendance or strategy change? Did you get fewer leads from events, or potentially fail to engage people at events you normally would have engaged?

Lead Quality

Lead quality. Are the people engaging the same quality? Sometimes as you scale up, quality goes down. That can have impacts lower in the funnel.

Lead timing. Are you reaching people at the right time? Some purchase windows are large–others are very small. You might be reaching the right people at the wrong time.

The purpose of marketing is to drive awareness of your solution—and, sometimes, the problem your solution solves if it's a new problem. Any change to your marketing top of funnel will cascade through the rest of your GTM.

The Lead Volume Problem

The most dangerous issue is not having enough leads. Not having enough leads creates problems whether you're a founder-led-sales organization of 2 people or a full-scale global sales team.

First, salespeople cling to leads and are reluctant to disqualify. This leads to inefficient sales processes and lower close rates.

Second, there are fewer chances to engage a lead and the perceived value of each lead is higher—resulting in less learning and salesperson (and process) growth, and taking fewer risks that sometimes lead to closed deals.

Finally, lack of leads crushes confidence. Nothing hurts a founder or salesperson's confidence like not closing deals–which is what happens when you don’t start with enough leads to begin with, you’re selling in fear of a “no”, and you don’t get enough chances to improve.

Some people don't like having too many leads. I disagree. Yes, there are ways to get a lot of bad leads—a TechCrunch article, a Product Hunt launch. These raise awareness and drive a spike in leads though most are not qualified or not actively buying. However, you can still engage the qualified ones. This is a great problem to have—far better than not enough leads. Just be sure your team is empowered to disqualify and move on.

Timing and Seasonality

Timing is crucial. For "pull products"—products where there's an existing market like HR software or accounting software—timing is everything. Are you reaching your customers too early, when they aren't in market yet? Or too late, when they've already picked a competitor and the switching costs aren't worth it?

Given a choice, better to reach them too early—but even better to reach them right when they're in a buying mood. See how AWS, Google Cloud, and others push their services through accelerators and startup products like Mercury and Brex. If they can reach a startup right when it starts developing, it'll be locked in—forever.

Recognize that for every channel there's some seasonality. For many industries, events are great in the fall and winter, but there are very few events in the summer or around major holidays. So if events drive your top of funnel and you're reading this in July, maybe that's what's happening. Same with paid advertising—it gets much more expensive over the holiday timeframe. So if your ROAS is down, perhaps higher costs are driven by more expensive ad rates across the entire industry.

If you discover that it's cyclical, make adjustments—but don't throw everything out.

Sales Top of Funnel

Not having enough leads from your marketing or outbound efforts can result in sales problems. But there are other failure modes for sales at the beginning of the customer journey.

For businesses that are relationship-driven and skip traditional marketing funnels, you might not notice a change in your marketing performance but still see a dip in your growth. Here are some areas to examine:

• Are your salespeople doing the same level of activity they were doing when things were going well? Same number of calls, emails, etc.

• Did the quality of your leads from marketing change?

• Did you see a change in high-quality, referred leads?

• Did anything change with your sales copy or messaging?

• Is your team responding to inbound leads quickly (ideally under 5 minutes)?

• Do you have nurturing set up and is it working properly?

Frequently, things change in the sales organization when a new sales leader arrives and shakes things up (accidentally or on purpose) or when a sales leader leaves. In a small company, that leader might not be a formal leader; they might just be a high-performing AE who comes up with effective ideas that others emulate.

A lot of issues on your sales team can be traced back to not having enough leads to begin with—so be careful here. Don't blame the sales team for problems that originate upstream.

Customer Journey Support

The customer doesn't know everything they need to know from one or two sales meetings or visits to your website. If you had PMF and a working GTM motion, then you should know what they do need to know in order to decide to purchase.

Look at how you support them and whether you removed anything that might have been a key pillar. This can happen during website redesigns, major rebrands, and at other strategic junctures.

If you never fully had PMF, but you're trying to figure out what's going wrong, then look at the journey for those who arrive in your top of funnel. Are they supported? If you were them, only seeing your company through their eyes, do they have enough information to decide to buy? Are you using language they use, and framing the problem in terms that they understand? Don’t talk about MCP and RAG… yet. I once had an SVP at a $25bn revenue company get upset at me for saying “API” in a meeting. True story.

A great way to understand what might be missing (and possibly even reinvigorate some sales discussions) is to reach out as a founder and ask if they can give you feedback on the process—especially the people who didn't convert. Were they the wrong people? Did they not have enough information to advocate for you? Did they go with another solution? If so, which one and why?

Closing, Pricing, and Negotiation

If people are getting through the journey and deciding to buy, things can still go wrong. Frequently, as we evolve our GTM, we might put different roadblocks, approvals, or other processes between a "yes" on a call and money in the bank. What does this part of the journey look like for your customers?

Here are some things to look at around closing that may be causing GTM problems:

Pricing. Is your pricing (still) reasonable and acceptable to your target customers? Sometimes new entrants, adjacent alternatives, or other market changes will make you lose PMF. Be especially careful after you launch any pricing changes.

Unnecessary delays. Are customers made to wait unnecessarily? As companies mature, more approvals are required for agreements, invoices, or purchase orders. All of these slow down your closing—and cost deals. Did something change recently?

Clear next steps. Are your salespeople asking for a clear next step? For unestablished companies without previous PMF, you may not be clear about what the next step is. Put in your credit card here, send you an agreement—what exactly are you asking your customer to do? Is it reasonable?

Legal agreements. Are your legal agreements reasonable and low-friction? I hate complex legal agreements. The chances that any agreement under $100k ever goes to lawyers is near zero. Act appropriately—you don't need the world in your agreements. Get deals done quickly. If multiple customers object to the same term, change it. The one exception here is credit card disputes—for low dollar values where you take Stripe, disputes are a real risk and having a clear agreement helps. But those sorts of customers are unlikely to redline your online terms.

Agreement sign-off time. The worst feeling for a salesperson is waiting for a signature from someone on their own team and worrying if the deal will fall through. Yes, it's good to have senior visibility into customer agreements. No, it's not good if this delays signing by days or even weeks. Better to have a standard agreement that's customer-friendly enough that you don't need senior or legal review prior to signing.

The closing process is still part of that customer's journey. If anything, it's the most important part, as it bleeds into being your "new user experience."

PriceFx made a big splash in the pricing industry by being easy to deal with—good form agreements with reasonable terms. This caused major pain for its competitor, a legacy company called Vendavo, which bordered on being antagonistic toward its future customers in its agreement negotiations.

Retention and Upsells

NRR is down market-wide. Just a fact. Jason Lemkin recently highlighted this in SaaStr, and I've seen it elsewhere—115% is now considered great, not the 130% of yesteryear. Companies are scrutinizing their SaaS spending more and more, given the overall uncertainty in the market.

Of course, issues can also be company-specific. I worked with a founder once who was sure that retention and upsells were going well. When we dug into his CRM, we found emails from customers to account managers where the customers asked to spend more with his company, and his team members were taking a week or more to reply. WTF? Why would you make them wait to spend more?

People-Led Retention

For people-led retention and upsells, responsiveness is key. Proactivity, also. Having account managers who are in touch with their customers is crucial. You can't just do this at renewal—it has to be an ongoing effort.

Things that can cause problems with GTM include:

Unresponsive account managers. Be careful that process or other improvements don't distract from the job of building relationships with customers.

Unhelpful account managers. Maybe you promoted people from customer success or other roles, but they aren't a fit for true account management. This is particularly critical in technical roles—be careful not to focus too much on sales in this role.

Market headwinds. Changes in the market can cause happy customers to renegotiate or cancel. Get ahead of this with great relationships.

Delivering on your promises. Did your product or service deliver on what was promised at the top of funnel? In your sales meetings? Churn can also happen when sales promised one thing but didn't deliver.

Product-Led Retention

For companies that upsell or retain in a PLG motion, where the upsells occur within the product, the issues can be different:

New competition. Did new competition enter your market? RB2B has a great top of funnel but 15% churn. I think that's because new players like Syft Data and Warmly (among others) are offering alternatives.

Adjacent market convergence. Did adjacent markets start to blend into your market? Warmly does a lot of different things, as does Apollo.

Customer changes. Did something change in your customers? Products that sell per-seat can be hard hit by layoffs and shrinking workforces. The shift to AI is also hurting some software companies, as is the shift away from climate spending.

In today's market with all the horizontal expansion, competition is becoming more fierce. Companies move into adjacent markets thinking they can expand their footprint with their existing customers. What they don't realize is that customer might have 2 or 3 vendors solving those adjacent jobs to be done who are also moving onto their turf—forcing a question: Keep things siloed or consolidate? Who wins that consolidation is a toss-up. It might not be you.

Debug Your GTM

Here's a thorough checklist to work through when debugging your GTM that I work through on engagements with founders here. You don’t have to be excellent at everything! If you are that’s probably a bad sign. Just be excellent at what has a high impact; it’s fine to be terrible at everything else. That’s why companies with, say, great word of mouth or relationship based sales (law firms, accountants, etc) have terrible websites.

The Bottom Line

The sooner you realize something is wrong with your GTM, the faster you can address it. At the core of any GTM process is the customer journey. The customer journey needs to be your starting point. From there, you can walk along the path to understand what your customer's experience is.

If you're a new company trying to figure out PMF and GTM fit for the first time, walk through the customer journey to find the rough spots and address them. For an established company where growth has slowed, look for what changed.

There's a difference between debugging and optimization. Optimization is focusing on what's working and making it work more. Debugging is focusing on what isn't working and fixing it. Both are important. But debugging is existential.

What's broken in your GTM right now? Reply if you’d like to discuss.

I was talking with the founder of Syft Data about marketing attribution. Imran has invested years in building a better way to identify leads both to act on them and to attribute marketing success.

I realized halfway through the conversation that attribution wasn’t the real problem. It’s just not that important. Of course, it’s really hard. I agree with that. But focusing on attribution is misleading.

Effective GTM has to be anchored in the actual customer's journey. Today those journeys are complicated. Understanding what led to a purchase seems like a critical input to making decisions about marketing–but that doesn’t mean focusing on attribution.

Attribution might be hard, but just because it's hard doesn't make it as important as some would have you think.

The hardest part of marketing

The hardest aspect of marketing–and the most important–is getting the attention of your customers. Whether that's for something entirely new, standing out as a differentiated product in a crowded field, or not standing out but having people pick your product even if you are undifferentiated, this is essential for building a company.

These are all different challenges and require very different approaches. OpenAI or Cluely are great examples of marketing something truly new and novel. OpenAI created a whole new way to interact with computers and solve problems. Cluely got its start when a college student built an app to cheat on an Amazon interview with hidden AI–and got the offer (but declined it). It has grown through viral, controversial marketing. Cluely overlays AI on your screen so that you can “cheat” at meetings, which to them simply means perform better. It’s a new use case–and needs brilliant marketing to succeed.

Skyp is an example of a product executing on a differentiation strategy in a large, crowded market. There are so many emailing tools I still learn about new ones–with tens of thousands of customers, doing $10mm or even $50mm in ARR. Our marketing is meant to define why a company should choose Skyp. In a big space, there's always room for one more competitor if it can differentiate.

Booking.com is one of the few tech companies that entered a market as an undifferentiated competitor on purpose and dominated it. Their thesis: nobody cares where they book–what matters is appearing high in SEO and having an optimized checkout flow. This creates a powerful growth loop in a competitive market (structurally lower CAC, more profit, poured back into more acquisition). The consumer doesn't realize they booked on booking.com until much later. They just clicked a well-placed ad and flowed through a streamlined checkout. For Booking's direct-response model, attribution is absolutely necessary–they need to know which ads work. But most B2B sales motions aren't simple funnels. Trying to simplify them into "Clicked ad → purchased" risks disaster.

When channels work, nobody really cares about attribution. Think about big law firms doing hundreds of millions in revenue per year. Do you think the partnership ever discussed last-click attribution? No. They talked about whether the TSR3 is better than the G430 driver. Because the clients just came. They were differentiated. Things were working.

What you don't do is sometimes more important

It is very tempting to focus on attribution as a problem worth solving. But what if you don’t? I would rather have an educated guess at what is working so I can spend my time doing more of it than waste time trying to be very, very sure what I thought was working is in fact working.

Attribution can lead you to bad decisions. At Flywheel, an early Uber competitor that sent taxis via app, a data analyst saw that we had a lot of riders starting rides at UCLA. The CEO thought we should target our marketing to UCLA. He attributed the growth in new users and rides as proof that our efforts there were working. We had to go where the attribution data showed things were working.

But I had serious doubts. I went to UCLA for business school. I know those customers. Even in business school we were all too poor to take taxis (and that was before Uber–which in LA was half the cost and 10x more convenient). When we dug in, it turned out that all of those rides–over 95%–were using a $20 promo code. That matched my understanding of the customer journey. This was not marketing working, even thought that’s what the analytics said–this was us giving away free taxi rides to students who would never actually pay for a ride.

Investing time in what is working is a way to establish attribution through experimentation. If you invest and it works even more–then you know. And you can keep investing. Whereas if you invested that energy into better attribution, you might know better whether it worked–but you won't have any growth to show for it.

I've advised many companies who over-tooled their GTM early. They have beautiful dashboards for execs to reference in meetings, but no high-functioning GTM channel. Unless a true direct response channel is the best channel (Google, in the case of Booking), even by your Series B, you're not going to know exactly what is working. And that’s fine. Don’t let that stop you from putting a very high confidence, polished GTM slide in your VC pitch deck–but it should stop you from focusing too much on attribution, rather than just doing what seems to be working.

When Series A or B companies get stuck, it's usually down to one of two issues. Either they over rotated on academic questions like attribution, at the direct expense of losing focus on doing things that worked. Or, a channel that used to work stopped working. Sometimes, both–a channel slows down, you dig in on attribution, and then that channel stops entirely.

At the beginning, attribution is easy

Attribution should reflect a deep understanding of your customers' journey. If you understand your customer's journey–and your GTM motions are working–then your attribution doesn't have to be perfect. You can ignore aspects of attribution because you simply know what's going on from talking to customers, observing them interact with your product, or other methods.

But before you can understand your customers' journey you need customers. Back in May, we didn't have anything-customers, traffic, nada. I wrote a guest post for Kyle Poyar's Growth Unhinged–a massive newsletter with 70,000 readers. That byline sent ~150+ waitlist signups to Skyp even though we didn't have a functioning product yet–and the article wasn't even about emailing!

Attribution was easy. We had 0 leads from any other source. The website was maybe 36 hours old. So every signup that week was, obviously, from Kyle's newsletter. Attribution got murkier as other newsletter collabs launched, I went on stage at SaaStr, SEO started to trickle in, etc.

Today we get ~100 visitors every day and I have no idea where almost all of them come from. And I'm fine with that. I'd rather have 100 visitors a day and no idea how I got them than 0 and be 100% certain.

This is the journey for all startups. Early on attribution is easy because the baseline is zero; any marketing you do must therefore get 100% of the credit–whether it's LinkedIn ads, a talk at a conference, or anything else. You can only do one or two things at a time, so it's pretty obvious what worked.

Talk to customers, not dashboards

As companies get larger, more and more of the team spends less and less time with customers. This is dangerous. And it can result in demand for more analytics. Rather than talking to or spending time with a customer, you look at the analytics in pursuit of understanding.

There are some amazing analytics packages out there. Posthog (and others, like Fullstory) let you watch session replays–actual videos of what the users actually did. Connecting GA4 to Google Search Console gives you decent (not great) data on what searches are driving users to your site. Some of the new LLM "AEO" or "GEO" tools like Profound can help you understand where you fit in ChatGPT, Claude and Gemini search results.

But nothing is a substitute for talking to customers and asking how they found you. Understand their journey. Whether that is simply an additional field on an onboarding form or a demo booking form ("How did you hear about us?") or something that you make sure to bring up in conversation, without it you are flying blind. Even if you have the best attribution tooling ever.

What we're seeing 4 months in

Skyp has been live for about 4 months and we're starting to see the customer journey play out. Some of our customers became customers pretty much immediately. We launched a self-serve checkout flow, and were shocked when someone used it to sign up the morning after his first call with us. That had never happened before–and it was pretty amazing.

The biggest change is that now we're seeing people come back who we talked to months ago. It makes attribution very difficult–but I'd rather be doing more demos and signing more customers than debating the accuracy of how those customers became customers.

Here are the top patterns we're seeing:

• Slow buyers. There are many who were part of that initial Kyle cohort who simply don't move quickly. Maybe they were in pain, but they don't buy the next morning. They're now coming back, booking demos, and converting.

• Didn't know they were in pain. We've been posting religiously to our Skyp Mode newsletter, a hyper-tactical take on outbound. People love it. They read it–open rates are even higher than this newsletter (which are already really, really high). They think they've got email figured out–but after reading Skyp Mode for a while, they realize maybe there are pains Skyp solves that they hadn't even thought of. They reach out for a demo (or just sign up).

• Refined ICP. Every new customer brings us closer to figuring out our ICP. What we thought it was when we started–early stage founders–still works. But we are seeing larger companies engage, and they have longer sales cycles than your 2-person AI startup. Attribution can’t tell you everything–even done really well. Like if you have the right people engaging with your content, marketing and sales.

• Reactivated people. I've been using Skyp to send reactivation campaigns to people who said they wanted a demo but never booked, booked a demo but no showed, signed up but never logged in, etc. Life happens. Sometimes they found another solution. Other times they were just busy with other priorities, and the reminder is enough to book that demo.

All of this is as valuable because it’s learning as it is because it’s working. It's rare that someone didn't buy from you but is still willing to share what they did buy and why. That's been happening a lot, and it's gold. Not only does it inform our product for features we're missing–but also it informs our marketing. Usually it's not a missing feature, but how we presented and packaged that lost the sale. The same for ICP–as people come back, we’re learning what their journey is so we can meet them where they are.

Feel bad for the big cos

Attribution gets harder for bigger companies. Big companies have been around for so long, and have so much happening, the reality of the customer journey gets lots in the noise. The well-tooled attribution funnel shows that a new signup clicked on a Google ad. But for all it knows, they were a customer 3 years ago at a prior job, and sought your solution out again. And just happened to click the “sponsored” link on the search results page.

The biggest startup companies–OpenAI, Anthropic–probably find it impossible to attribute most of their signups. A friend at work. A newsletter. A TV ad. A post on social. Who knows?

Who cares? As long as it’s working.

Yes there comes a point where you should invest in attribution and analysis–but that point is much later than most founders think. And it’ll never tell you everything. That never means you take resources away from doing what’s working. That’s a failure mode.

So what should you focus on now?

If not attribution, then what? Three things:

• Customer conversations. Make "How did you hear about us?" a standard question. Not just for attribution data, but to understand their actual journey. What were they searching for? What problem were they trying to solve? How long had they been looking? If you optimize for this in your content, you’ll also get more LLM traffic. A good friend is CEO of an insurtech AI startup that is doing exceptionally well and has raised over $50mm–and he still flys around the country meeting with customers every week.

• Rapid experimentation. Try things. If you invest more time or resources into a channel and it works even more–then you know. Might not play well in a board deck, but if you show your board consistent growth, you'll get fewer questions about attribution to begin with.

• Double down on what resonates. You'll feel it when something works. More inbound. More engaged conversations. More people saying "this is exactly what I need." Do more of that, even if you can't perfectly measure why it's working.

GTM is changing faster than ever before. The channel that works for you in the early stages won't be in any attribution dashboard. Don’t worry about perfecting your attribution–just go figure out what works. You’ll know.

Welcome to Seed to Sequoia, the newsletter for founders about GTM and strategy. Looking forward to TechCrunch Disrupt next week–if you’ll be in town, let me know! Would be happy to catch up f2f IRL–grab a time here.

One ask: what (other) newsletters do you read and enjoy? Just reply…

This post is brought to you by PicoMES. The next generation of manufacturers are building smarter, faster, and with fewer mistakes—and PicoMES is their partner on the shop floor. PicoMES helps innovative factories digitize and streamline assembly operations with modular software that scales as you grow. From digital worker guidance to error-proofed workflows with full traceability, companies as small as 6 operators to those as large as 1800 per shift trust PicoMES to bring startup speed and enterprise-grade reliability to modern manufacturing.

Every founder has obsessed about something.

For me it was a guy named Will.

Will had been VP Sales at a direct competitor that went public. He built the Europe business from $0 to $50mm in ARR and was part of the leadership team when the company went public. He reached out one day because he heard about us. I showed him a demo. He thought we had the Best Thing Ever. Immediately he wanted to sell it.

There's nothing as validating as hearing someone deeply experienced in your market get excited about your product.

After a couple of months of getting to know each other, we hired him. It was a huge step to us as founders–particularly to my cofounder. We had a bona fide VP Sales who was going to get us to millions in ARR–and quickly. It felt like an inflection point–we were about to achieve escape velocity. We had funding for it–but it was a stretch, and more than we were planning to spend on sales at the time.

I'll skip to the end: six months after starting, he had only closed one deal–that I had gotten to the 1 yard line. We had flown all over the world, spending 10x more on T&E than before, meeting with several totally unqualified (but great-sounding) prospects who had no intention of buying. It clearly wasn't working.

Frustrated with the lack of success, when a pilot that seemed to have gone well didn't convert he was incredulous. He threatened to take their data and sell our solution to a direct competitor if they were "too stupid" to buy. That was (obviously) the end of things. I forgave him–the botched pilot wasn't only his fault. But we all felt the pain, no one more so than me. Even after we parted ways, we never got that customer back. Some damage is permanent.

How magic bullets kill companies

Whether it's a key hire, a shiny new agency, a partnership, or a top-tier VC, founders obsess. They buy into the marketing, the branding, the reputation. It's hardwired into being human. But these rarely work out the way founders hope.

In reality, hiring the wrong person can be very damaging. But hiring the right person is just one step on the path. It's not a magic bullet.

Thinking you have the magic bullet can kill companies. One company I worked with hired a CRO, thinking this would solve all of their sales problems. That CRO turned out to be terrible. Good people left, the sales team he hired (for ~$1mm/ year) never sold a single dollar of software; renewals ground to a halt as he broke the part of the sales motion that was working, while what he started never worked. And in the background he spent $500k on a botched migration from Salesforce to Hubspot.

Recently a founder explained why they chose a Tier 1 VC–because of their ability to help with recruiting. After the investment closed, he reached out to their recruiting partner–to learn they had a 9 month backlog.

Another company took a seed investment from Sequoia Capital, the most famous VC firm. They promised the data the startup needed to prove their algorithm worked. Their portfolio companies would happily share it if they asked, just sign the Series A term sheet and LFG.

Thing is–the portfolio companies never shared that data. Sequoia won a competitive deal, on very investor-friendly terms, but never came through on that key deal point. They hired me to go get that data–and I did. To be fair–I wouldn’t have considered working there if Jim from Sequoia hadn’t been on the board. But that wasn’t the magic bullet the founder thought she was getting.

How to recognize magic bullet thinking

Founders need to take risks. That's the whole point of the exercise. If Zuckerberg hadn't taken a chance on Sheryl Sandberg, Facebook would look very different–and only in a worse way. But it's very easy to overestimate the expected return of any given risk.

Some of them will pay off big time. Many will not. And almost none will pay off immediately–so be aware.

Here are five ways I check myself for magic bullet thinking.

One way doors

Magic bullets usually present themselves as one way doors. Jeff Bezos popularized the idea of one way and two way doors in his 1997 shareholder letter. A one way door is "consequential and irreversible". He explained that these "Type 1" decisions are heavy, and should be done thoughtfully and slowly, while "Type 2" decisions are like two way doors, which you can simply undo. Type 2 decisions should be made quickly by high judgement individuals or small groups.

Magic bullets have a tendency to look like Type 1 decisions. A major software purchase, a key hire with a large salary and equity grant, a 12-month agency contract, hiring an investment banker (with a $500k minimum and 12-month tail). Sometimes the very people involved will make you feel like they are a Type 2 decision–no big deal–when they are, in fact, a very big deal.

Retaining that investment banker, for example, is a Type 1 decision. They will run your transaction and can very possibly ruin it. Their presence on smaller transactions can scare off buyers. And if you don't get along, or they don't do a good job, you're stuck with them for a year–which might be all the runway you have, anyway.

They'll present this as an easy decision. Their agreement is often called an "engagement letter"—which sounds lighter than it is. But it can very much be a one way door.

Incremental proof

Can you turn your magic bullet thinking into an incremental experiment? This will tear you away from one way door thinking, and help you pin less hope on a magic bullet and be more rational.

For example, could that VP hire (always a one way door) be an advisor for a few months first? Has that investment banker provided any value prior to negotiating their engagement letter? A banker who has made introductions to possible acquiring CEOs and has back channeled for you already is much less risky than one who essentially responded to an RFP.

If you cannot turn it into an incremental experiment, then you are very likely dealing with a magic bullet. That fractional CRO or agency that requires an annual contract. That AI SDR software that requires an annual contract–but will solve all of your top of funnel problems like AI magic.

While there might be rational reasons for these terms, often it's because the vendor wants to lock you in before you realize the results aren't magical.

What has to be true

You're making assumptions–which is fine; without them you'd be paralyzed. But it helps to make those assumptions explicit.

If you hire this CRO, what has to be true for that to have been a good decision? For example: they need to be good at working in an environment like yours. Your sales cycle and process should be similar to their last company. Your top of funnel needs to create enough demand for them to hire the sales team they said they want to hire. Your marketing team has the infrastructure and collateral they'll require to succeed. Your demos convert to paid customers. And so on.

How many of these have you proven and how many are assumptions? The more unproven assumptions, the more likely you are thinking that this hire is a magic bullet.

Another way to think about what has to be true is to pose the opposite question: could the opposite be true?

If you're thinking "I have to hire a CRO right now", what if the opposite is true: "I absolutely should not hire a CRO right now". What would have to be true for that to be the case?

What if it fails

If you write down a failure post-mortem for six or twelve months out, what happened? You're not necessarily looking at what led to the failure (though that can help) but rather what the impact is.

How bad is that impact? We had $10 million in the bank, now we have $3 million and only 4 months of runway. Our return on ad spend was 1.6 but the new agency did a bad job and it's now below 1 (that is, unprofitable). That new CRO angered our biggest partner and we lost 1/3 of our pipeline. Any of these scenarios might be company-ending.

If the risk is existential for your company, or could cost it a year or more of progress, then it is probably magic bullet thinking.

Why now

People tend to fixate on the things that are urgent, especially if they seem important. For example, a VP candidate has an offer and needs to decide within a week. She's pretty good, maybe not perfect, but you need to decide ASAP. This kind of urgency–real or false–leads to magic bullet thinking.

Vendors will create false urgency. Is a 20% discount for an annual commitment for software you haven't tried and aren't sure will actually solve your problems a good idea?

Probably not.

But when a salesperson offers that deal and your team has 2 days before the end of the month to decide–some will see a magic bullet and say yes.

Expertise asymmetry

It's hard for salespeople to hire engineers. It's also hard for engineers to hire salespeople. If you are hiring for an expertise that you yourself do not possess, recognize that you are not capable of making that assessment–and you might be seeing magic bullets.

A "great ML engineer" who costs $1 million a year might seem like the solution to all of your problems–but if you don't also have a PhD in machine learning, how can you know they have the skills to solve your specific problems?

The same is true of nontechnical roles–sales people, marketers, agencies, lawyers. I've seen founders hire "experienced enterprise sales leaders" who turned out to be good at interviewing but bad at selling—and the founder couldn't tell the difference until a year of runway was gone.

A friend calls these people names like "Product Jesus" or "Sales Jesus" because they'll supposedly save your struggling product or failing go-to-market. The more messianic the promise, the more careful you should be. Nobody walks on water–and if someone claims they can, or you invent in your head that they will–it’s a magic bullet.

Be careful and get advice from founders or investors when hiring for expertise that you don't possess. The worst outcomes come from hiring an expert who is not actually an expert, listening to them (like you're supposed to), and then failing as a result.

A real, recent example

A founder I advise was excited about hiring a new agency—so excited he called asking for fundraising advice to afford it. It was urgent.

Except the urgency was all in his mind.

He'd gotten a meeting with the top marketing agency in his space. They run campaigns for all the now-household names. They're legit. And they have a great vision for his brand.

But they cost 4x more than his current agency, require significant onboarding, and a long minimum contract—forcing him to raise money as debt or equity.

He was acting like he'd switch within days, even though he'd only met them earlier that week. I asked: Why the rush?

When he stepped back, he realized he was chasing a magic bullet. His current agency is doing exceptionally well—and he's their best client, so they jump to fix any gaps. If you go from being a small agency's best client to a big agency's smallest client, could results actually get worse?

Yes. Yes they could.

Did he want to raise money now? No. So why sign an agency that required it?

What if he just took that incremental spend and put it toward more media with his current agency? Lower risk, likely higher return, and he could always switch to the bigger agency later.

He tapped the brakes. Wisely so.

Be the magic bullet

Let's reverse all of this. For your customers, you want to be that magic bullet. The one thing that if they do will solve all of their most important problems.

Salesforce will make you a well-organized company with rapidly growing sales—a trailblazer.

Hubspot will bring you leads. Lots of leads.

Apollo will get your emails answered and fix all of your top of funnel problems (not as well as Skyp, but you knew I'd say that…)

The more you can sell the vision and fit into your prospect's mental model of what that magic bullet is, the more likely they are to set aside their doubts and put in their credit card.

Be bullet proof

As you run your company, be aware of magic bullets. They'll seem like normal, sound strategic or tactical choices. But if you're not thinking of any downsides, If you're not asking yourself if the opposite could be true–tap the brakes.

When something seems so good you can't think how it wouldn't work–you might be being looking at a magic bullet. Check references, do your research, and make sure your expectations fit with reality.

Sure, act decisively-maybe it'll work out. Sometimes it does. But be self aware, and don't bet your company on it.

Instead, become that magic bullet for your customers. And if you deliver–even better.

Welcome to Seed to Sequoia, the newsletter for founders about GTM and strategy. We had a huge turnout for our #SFTechWeek event–430 signups and over 200 attendees! It was so big that we’re doing it again–online. Want to hear how to get your company acquired? Register now!

This post is brought to you by PicoMES. The next generation of manufacturers are building smarter, faster, and with fewer mistakes—and PicoMES is their partner on the shop floor. PicoMES helps innovative factories digitize and streamline assembly operations with modular software that scales as you grow. From digital worker guidance to error-proofed workflows with full traceability, companies as small as 6 operators to those as large as 1800 per shift trust PicoMES to bring startup speed and enterprise-grade reliability to modern manufacturing.

GTM software unicorns are struggling. Everyone is expanding into others’ turf, increasing costs while competition is hurting margins. You know the names:

Gong. Outreach. Apollo. SalesLoft.

It’s critical–as always–to have a real product philosophy underpinning product decisions. Sure, you could build a LinkedIn chrome plugin. But should you? Here’s one way of thinking about when to stay in your lane, and when to swerve.

Knowing your lane starts with your customer

First you have to know what your lane is, and clearly communicate that to your team and ecosystem. If you’re very close to your customer’s needs, it’s easier to decide what is and isn’t out of scope.

If your customer struggles with something that is core to the value you deliver, absolutely absorb it into your product. But if you are expanding just to get more TAM, just to grab more pie, think hard before you do.

Never let your competition dictate what you build. Instead focus on your customer. If competition is doing something your customer demands, and you are not doing it, then you need to step up. But if you’re just copying your competition, it can get ugly.

Apollo started as a database to find contact information to power outreach. Then it added the ability to build email sequences, because (obviously) that’s what everyone was doing with the data. Making it easier to send sequences made it easier to sell more credits for contacts.

Of course, in response all of its competitors–the email sequencing companies like Lemlist–added databases and the ability to search for contacts. I think this was probably inevitable–they are separate jobs to be done, but very closely related. But I’m sure as one company stepped onto the others’ turf, the other got up and did something about it. Now 2x more development is required to do basically the same thing. That can’t be good. Apollo’s email sequencing tools were never very good. Yes, they looked great, but their deliverability has always been very poor and usability leaves a lot to be desired.

42floors tried something similar–and had to fire half its team and roll it back. 42floors was a top YC company that was like Zillow for commercial real estate. It scraped all the commercial listings, eventually attracting a huge percentage of those listings and agents to its platform (as advertisers). If you Googled anything related to commercial real estate you didn’t find much–until 42floors, which owned the search results similarly to how Zillow owned residential listings.

The team thought that there was more money in being an agency, rather than just an advertising website (think Redfin instead of Zillow). They hired agents and got the regulatory licenses to act as an agent. Immediately every listing fled their platform. The other side of the marketplace wasn’t ok with competing with the platform. Jason Freedman had to let the entire brokerage team go and roll the whole program back.

As all of these applications expand into each others’ turf, I wonder how much the additional stakeholders are reacting. For example Lemlist I’m sure referred people to Apollo to find leads. But as soon as Apollo built email sequencing, I doubt lemlist continued to send people that way. Instead it built its own. Eventually this has to impact CAC and pipeline.

Why you have to expand

There are three reasons why you have to swerve beyond your lane. They are all defensive. Competitors are moving in, or could move in, from adjacent spaces. Your lane cannot support a viable business. Or your customers demand you do something different or that you don’t currently do. Here are examples of each.

Facebook got into messaging by buying WhatsApp because it realized that it had to. Its users were spending more and more of their time in messaging apps, especially that one. It did the same with most other communications-related innovations, including a Tik Tok ripoff within Instagram. I would argue this is only kind of leaving their lane–their lane is communications. So, maybe they didn’t leave. But they sure expanded their product footprint.

The classic reason involves leaving your lane because it cannot support a viable business–but an adjacent space could. There are countless stories here. One that I was very close to was TubeMogul. They started as a consolidated analytics platform for online video. Before YouTube dominated, there were a dozen online video outlets and creators–before anyone called them that–would post to every one. Aggregating stats was difficult, as was actually posting to all of those sites. TubeMogul was software that did that for people, and charged a little for it.

TubeMogul had built a large stable of customers, but realized all the money was in the ads. So Brett Wilson and team kept the tech to post and manage analytics, but built an ad network and started selling ads on all of those creators’ channels. When TubeMogul went public it was one of the largest online video ad networks.

When customers demand something, it’s an easy choice. Hubspot got into CRM because it was an obvious extension of inbound marketing. You’ve built a blog, landing pages, all these lead capture forms… and then what? You have to try to deal with them in Salesforce? Customers really wanted a sequencing tool, so they started there. If it hurt their relationship with Salesforce, so be it.

When not to shift lanes

You have to establish yourself first before you can start expanding out of your lane. All of the companies mentioned above were established, successful businesses before they expanded into something else. Early stage founders (myself included) like to shift to the next shiny object. But be careful not to build just for building’s sake. Sometimes you do need to fill in a core missing piece. But make sure you’re doing it because the customers asked for it. Not to fulfill the whims of non customers–investors, future acquirers, or even founders who think it’d be a cool thing to build.

Distraction vs. vision

Google launched with the idea of “organizing the world’s information”. This guiding purpose underpinned everything from building a search engine to building a web browser to building robots to illegally copy books at massive scale (and get away with it). Having that core purpose makes it easier to determine what is your lane and what isn’t.

Early on your lane should be exceptionally narrow. Anything else is a distraction. Google’s lane was just the search page. That’s it. No extras. Later they earned the right to shift into other lines of business. Even though to this day search remains the majority of its revenue (~58%).

There are exceptions that people who want to veer constantly bring up. Hubspot has all kinds of features–but that is kind of their lane. Integrated (but never best-in-class) marketing tools. PostHog, which gives a ton away for free and seems to do everything. These are the exceptions. For every one of them there are far more successes–Tally, for example–that stayed in their lane.

Stay in your lane–swerve when you have to

For most startup founders, the best advice is to stay in your lane. Make sure you’ve really nailed the jobs to be done for your core customer. Then you can look around, and ask if there’s maybe a more valuable problem you could tackle for them. At this point you should have distribution figured out, so the pivot or expansion is far less risky. If you swerve too early, you’ll crash.

I said I’d dig in on LinkedIn–here’s what I found. Hopefully you are crushing LinkedIn and getting a lot out of it. If not, or if you’re still ambivalent, perhaps this post has some utility. If I missed anything, or you have comments–which I expect many will, as many of you have 10k+ followings–please reply! Would love to hear about it, and I’m sure many would appreciate a follow up post.

This post is brought to you by PicoMES. The next generation of manufacturers are building smarter, faster, and with fewer mistakes—and PicoMES is their partner on the shop floor. PicoMES helps innovative factories digitize and streamline assembly operations with modular software that scales as you grow. From digital worker guidance to error-proofed workflows with full traceability, companies as small as 6 operators to those as large as 1800 per shift trust PicoMES to bring startup speed and enterprise-grade reliability to modern manufacturing.

You spend how much on LinkedIn content?!

You don’t give 2 shits about impressions?

You only spend an hour or two a week and get 1mm impressions?

I was shocked when I reached out to people in my network about LinkedIn. I’ve had a love-hate relationship with LinkedIn, and I doubt I’m the only one.

I came back from vacation this summer to crickets–posts with 200 views where I had been getting tens of thousands. Having invested time and energy over the last 12-18 months, I had questions. Was this a good use of my time? Am I doing it wrong?

So I started asking around. Of course, when I posed this question as a post–and it didn’t get any reach. But I reached out in DMs and email and got some very interesting replies. The friends who responded have a combined ~400k followers, which I only mention because it’s a great vanity metric 😎.

I asked just three simple questions:

• what metrics you measure and what are they (ballpark is fine); eg impressions per post or week, followers, follower growth, meetings booked etc.

• how much effort you honestly put in, measured however you want (time, money, team, both)–again ballpark is fine;

• what if anything has changed from a performance standpoint and/or what changed in the past and how did you adapt. 

There are many paths to success on LinkedIn, and it surprised me to hear how vastly different approaches led to the same place. The other surprise was that when I subconsciously did some of the things these conversations led me to, I had some success with this post.

This is not a deep dive on how to excel on LinkedIn. If you want that, Kyle Poyar wrote a great Guide to mastering LinkedIn. His input is in this article. Highly recommend.

This not a deep data piece in which I crunched a ton of data on millions of posts or whatever. I’ll leave that to James He, whose company Artificial Societies aims to simulate and improve LinkedIn engagement. His input is also included.

Instead I’m just looking to give you an idea of what great looks like, how to measure it, and how to figure out if you’re on the way. Or to give you permission to throw in the towel, and spend your time on something else.

The metrics that matter on LinkedIn

Everyone pointed to similar metrics. The most commonly tracked are:

• Impressions

• Engagement (comments, reactions)

• Reposts

• DMs

• Meetings booked

Impressions are extremely noisy, and almost nobody cares about them as a primary metric–but most keep an eye on them, and know them off the top of their head. Adam Robinson recently posted that he gets ~1 million impressions a month on LinkedIn. But he also gets insane engagement, which is more important.

Engagement–likes, reposts, comments–is a far better measure of whether a post was well received. It also can give you something to follow up on. Unfortunately, some of this can happen off platform–but it’s still great. For example, a former FTC commissioner sent me a text about this post. His current employer forbids commenting on LinkedIn.

What happens when you put the FTC at #7 on things that kill M&A deals

Engagement is not highly correlated with impressions. For example I was talking with someone who had a post that had 100+ likes but 15,000 impressions, where the recent, successful post I mentioned above had 50 likes and 44,000 impressions.

Engagement is also better because it is actionable. You can see who is engaging with which posts. Tools like syftdata.com automate reaching out to people who engage with your posts and are in your target ICP. An impression gives you nothing to work with (except, the numbers are usually in the thousands, which is cooler).

Some engagement is better than others. Reposts will expand your reach more than comments, but comments will expand it more than reactions. Lots of reposts usually signal that that post hit. (In the parlance of our times.)

To get better engagement, everyone said the same thing: focus on excellent content that generates reaction or conversations. Obviously, but also – not obviously. Because we all post things that don’t get reactions or generate conversations, but (if you’re me) you keep posting stuff like it.

Just. Stop.

Nobody I talked to uses AI on its own to generate posts (though James, obviously, uses AI to refine what he posts, and he’s not the only one). Instead they are writing posts that are insightful, contrarian, or draw on data. Most use AI in the either the drafting or editing process, but nobody has gone fully AGI on their LinkedIn post generation.

Great posts don’t have to be original or original research. For example, that 9-9-6 post originated from something Peter Walker posted, which originated with a graphic someone else posted first. But it hit a nerve.

Community driven distribution

Artificial Societies has a great feature where it tells you who your post resonated with and why. On one post I ran through it, it concluded that founders would engage but VC’s wouldn’t because something it said was too honest and even though they likely shared the sentiment, their LPs might get concerned 🙂. Considering how your audience will react–and whether they can react, without repercussions–is important.

Writing for a specific audience on specific topics is the best way to resonate. There are things I write that I’m sure no 9-to-5 employee at a major corporation would click “like” on. But founders, VCs, other startup world people do.

Contrarian posts do well. This is a fact. They can still do poorly, but I think the reason why they are more likely to do well is because they are controversial and start a conversation. If you know your community and can write controversial things that maintain your reputation in a positive way, that can be a great way to generate more engagement.

Consistency in – but not consistency out

Everyone I talked to posts regularly. Some post 2-3x a week, some post every day.

Nobody I talked to gets consistent results.

Even people at the top of their game have dud posts. Their dud posts might do better than my best posts, but they are still relative duds. There is not much correlation between time invested in a post and how well it does on the metrics. For me, my best performing posts are almost always hot takes that I wrote and sent in <10 minutes, before I gave it much thought (if I had, maybe I wouldn’t have posted!).

It seemed like the people who had a team or were part of a team experienced more consistent results. Two had similar processes. They set aside half a day per week to iterate on post ideas and draft posts with his team. One would then engage with others on the platform a little each day, but the drafting and posts was a contained exercise (with a team).

That is not to say the solo creators were less successful–they were among the strongest performers overall. Perhaps they had more self discipline, or perhaps they had their own process but didn’t have to document it or think about it, because they were just doing it. I got the sense that the solo creators all really enjoyed LinkedIn so it didn’t feel as much like work as it might if you were, say, a founder/CEO with a bunch of other stuff on your to-do list. People with a team seemed to be doing it for business purposes and were intentionally efficient.

Effort varies widely

Everyone I spoke with had well over 10k followers–some had 100k or more. But how they got there and what they do now that they’ve got a significant audience were very different.

There were three categories of effort:

1. Just post stuff that seems interesting to me, 3-20 hrs/wk

2. Do deep research (possibly with a team) and post the analysis (10-20+ hrs/wk)

3. Hire a team (in one case, $25k/mo) to do the stuff

You do you, I guess? I really don’t know what to make of this. Obviously if you’re paying for a team–you should expect results, and they should expect to have to deliver. There are teams out there who can do this for you successfully.

But if you don’t have $5-25k/mo to spend on professionals, what you invest in time is not necessarily correlated to success. I’ve seen this myself, with some of my best posts being short, quick, not terribly well considered. Some very thoughtful posts also did well, but there’s little correlation between effort and results.

Some people just like spending time on the platform, and in doing so improve their reach. Commenting and reacting to other posts draws people to their profile, including their own network. That leads to better performing posts, especially if you engage after you do a post (people will find it).

Results

I’ll use impressions for comparison because most people referred to their impressions when talking of overall results–ironic because everyone said engagement is more important! I’ll break it into large followings and smaller followings.

• Large followings (>50k)

  • ~1mm impressions a week.

  • Growing ~1-4k followers per month

• Midsize Followings (10k-50k)

  • 50-100k impressions/wk

  • 5-10% monthly follower growth = 500-1k/mo

Yes, I should add, people do track follower growth. But it doesn’t seem to be a big focus for anyone, just something to keep an eye on.

Even people with midsize followings could get 300-500 engagements. That’s a lot of possible leads to sort through! Maybe use AI?

Many people I spoke with track meetings booked or new customers or newsletter subscribers. It’s become very hard to track website source traffic accurately; nobody mentioned it. But one person with ~10k followers was booking several meetings a week, even without viral posts, using the “book a meeting” call to action that LinkedIn enables in premium profiles. He was simply reaching his target audience consistently.

Before you go set your profile to the “book a meeting” call to action, realize that this can also go badly. We posted a job at Skyp and I had half a dozen meetings a week on my calendar that were recruiters trying to get my attention.

The term “social selling” refers to not only posting to LinkedIn (or other social networks) but also following up on that platform. If you’re getting 300 engagements a week, you (or someone on your team) need to look at all 300 people, find your ICPs, and message or interact with them in some way. Possibly, off-platform (using an enrichment tool like Apollo to get their email address and email them, for example). You will get more meetings this way than if you just hope people click a meeting link.

The playbook

Here’s what worked across everyone I talked to. No huge secrets in here, but it’s useful to hear the same things repeated by several people who are succeeding.

• Prioritize engagement-driving content → Ask provocative questions, share tactical frameworks relevant to your audience, maybe post Looms/videos for depth.

• Anchor in community → Content should feel like it’s written for insiders.

• Consistency → Maintain a reliable cadence of posts and engage regularly with others on the platform.

• Engage also → Just posting and walking away is not the way. Engage with commenters on your posts, and engage with other influencers posts who have audiences that are the audience you want.

• Boost → Most people are using boosts to get more out-of-network reach for posts that were already successful. LinkedIn just made this easier.

• Hire experts if this is mission critical → If succeeding on LinkedIn is crucial to you, and you have the scratch, why not? If I had to work with someone, it would probably be Finn McKenty. If I were bringing on a content team, I’d talk to Adam at KnownFor.

Is this your experience?

I chatted with Finn recently about my frustration and how I felt about wasting time. He gave me permission to just stop doing LinkedIn, and that felt great. I also recently added a “how did you find us?” to the questions asked when a person books a meeting with me in Calendly. Guess what the last meeting said?

“LinkedIn”

So maybe I can’t quit yet?

I’m curious if this reflects your experience. Just reply!

This week’s guest post breaks our pricing streak with a post about disruption. Danny Nathan writes a great newsletter and podcast called Innovate, Disrupt or Die where he focuses on topics relevant to founders. I invited him on to talk about his bread and butter, disruption. Danny founded and runs Apollo 21, a design and venture studio. You can find him on LinkedIn.  

This subject rings particularly true for me because of Skyp.ai. We’re serving an underserved market. Some have questioned why I’m building yet another email product–after all, Apollo has 500,000 customers. And that’s exactly why–our less-full-featured, bootstrapped product is winning customers away from them. Just like some of the examples Danny talks about, it’s not about competing head on. 

This post is brought to you by PicoMES. The next generation of manufacturers are building smarter, faster, and with fewer mistakes—and PicoMES is their partner on the shop floor. PicoMES helps innovative factories digitize and streamline assembly operations with modular software that scales as you grow. From digital worker guidance to error-proofed workflows with full traceability, companies as small as 6 operators to those as large as 1800 per shift trust PicoMES to bring startup speed and enterprise-grade reliability to modern manufacturing.

Why Aim for Disruption?

Most founders, or innovators for that matter, don’t set out to disrupt an industry. They start with a problem worth solving, a job to be done. The moment of disruption shows up later–if it shows up at all. 

Strategic disruption unlocks new markets by surfacing customers and use cases incumbents ignore. Done right, disruptive growth isn’t dependent on stealing market share or defending margin. It comes from offering something the existing system wasn’t built to support and doesn’t care to challenge.

All of which begs the question: can disruption happen deliberately?

The Path to Deliberate Disruption

Disruption works because it makes room for something cheaper, easier to use, or more accessible. If you can surface those opportunities early, you can design your solution to them, on purpose, from the get-go. The outcome isn’t a side effect. It’s a design choice. If you want to open new markets and unlock runaway growth, you can’t just hope to stumble into it. 

You have to engineer your way there.

Planning for disruption begins with mapping where entrenched systems are weakest, then filling in the gaps others can’t (or won’t) address. The most effective disruptions begin with a clear line of inquiry… 

What parts of this system are treated as immutable?

This work looks unglamorous at first. You’ll research deeply. You’ll strip features. You’ll explore pricing that finance hates. You’ll route around channels that once felt sacred. 

The friction shows you it’s working. If you’re not challenging those defaults, you’re not architecting anything meaningfully different or disruptive.

Misconceptions and Missteps

“Disruption” sounds like conflict, so people assume it will show up as a David vs. Goliath battle, ultimately resulting in a startup crushing a market leader.

While that makes a great story, it’s the wrong strategic approach. Disruption rarely looks impressive at first. That’s the point. It sneaks in by serving the needs incumbents ignore because meeting those needs would cannibalize margins or threaten positioning.

New Value Networks > New Technologies

Disruptors don’t win because they have better technology. If you’re building with cutting‑edge tools but relying on legacy assumptions, same pricing, same channels, same margins, you’re not disrupting. You’re modernizing. 

“Shitty” products, aimed at overlooked problems and underserved customers, can become a wedge into massive markets. The advantage isn’t technical superiority, it’s fit and accessibility. New value networks emerge when you shift power. Technology matters, but only as a means to reshape the architecture of the value networks you’re operating in.

Most products that are ultimately seen as disruptive begin in low-end or new-market footholds with a feature set that is cheaper, simpler, and just good enough. And that’s why they’re often dismissed by incumbent leaders as unimportant or unprofitable. 

That dismissal is the opening.

Disruptors win because they build systems that behave differently. Systems that serve different people, create value in new ways, and challenge who benefits. Technology is simply a tool to build those systems.

Disruption in Action

During a recent podcast debate, a guest said, “I hate the word disruption. Canva didn’t disrupt Adobe. They just found 170 million people Adobe wasn’t talking to.”

He’s right. Canva didn’t replace Adobe.

But Canva did disrupt the assumptions the entire design‑software industry had been built on:

• Design is for trained professionals

• The average person doesn’t need to design anything

• Design tools should be expensive and complex

These beliefs shaped the way the industry was built. They informed how companies designed products, priced access, and marketed capabilities. And for years, nobody argued when an Adobe Suite license cost $600+ per year.

When Canva came in, they didn’t abide by those assumptions.
They didn’t compete on those terms.
They redefined both.

You Know What Happens When You “Assume”

The same rules appeared everywhere: professional tools must be powerful rather than simple, real design requires training, and the user base is a small, high‑value segment. When Canva entered the market, they dismantled these underlying assumptions about who design software is created for. 

And, in doing so, they made Adobe’s framing of the market obsolete. 

Canva proved there were millions of people who wanted to design but didn’t consider Adobe’s tools an option. Not because those people weren’t aware of Adobe products, but because they were never invited into the fold.

When Canva offered a tool that said, “No experience needed, start now, it’s free,” they didn’t just offer a product. They rewrote the rules. And opened a massive segment of “non‑professionals” the industry had ignored.

If that sounds like dumbing things down, look closer. Canva made a choice to optimize for reach, speed, and social collaboration, not for expert control. That choice changes who creates, who pays, and where the growth comes from.

To Disrupt, Question the Rules Others Don’t

Today’s founders can take a cue from Canva’s approach. Before you interview customers or write a roadmap, map the assumptions that quietly define your market. Treat this like a forensic exercise. 

• Ask how pricing is structured and which margin rules make segments look unattractive. 

• Ask who holds power in the value network, where the gatekeepers sit, and what the switching costs are.

• Ask what qualifies someone as a customer and who gets excluded by that definition. 

• Ask which “minimum” features actually drive cost and complexity and what can be stripped without killing the result. 

• Ask whether distribution models or technologies changed in ways the category ignored, and which new channels lower access friction.

This is where new markets emerge: from people who’ve been excluded or underserved because they didn’t fit the old logic. Once you find those people, talk to them. Understand the jobs they’re trying to get done. Watch how they solve the problem today with hacks and workarounds. Let their constraints shape what you build, how you price it, and how you communicate value.

These core assumptions are often the most fragile, and the least questioned, parts of the system. Interrogate them deeply from day one and keep questioning them throughout your customer development process.

Case Study: How Nintendo Redefined the Rules of Gaming by Ignoring the Market’s Assumptions

When Nintendo launched the Wii in 2006, they didn’t just ship a new console. They rewrote the logic of an entire industry. At a time when Sony and Microsoft were locked in a race to push better graphics, more power, and higher performance, Nintendo walked away from the arms race entirely. 

They bet on movement. On simplicity. On fun.

That wasn’t a fluke. It was a deliberate act of disruption. Nintendo questioned the very assumptions that defined gaming at the time: 

• Success required hyper-realistic visuals

• Serious gamers were the only market worth serving

• Performance would always beat accessibility. 

They chose different customers, different hardware, and a different definition of value.

The Wii was underpowered by industry standards. But that was the point. It was easier to use, cheaper to produce, and radically inclusive. Nintendo didn’t win by outperforming Sony and Microsoft. They won by targeting people their competitors didn’t even acknowledge: families, seniors, and first-time players.

Playing Wii. Image Credit

The results were undeniable. The Wii sold over 100 million units, outselling both the Xbox 360 and PlayStation 3. In just one generation, Nintendo went from being perceived as the quaint third player in the console race to owning the top spot globally. By the end of 2007, the Wii was outselling both competitors in every major market. And over 60% of its players were new to gaming entirely. Nintendo didn’t just compete better. It expanded the category.

Disruption wasn’t a side effect. It was the result of challenging assumptions others treated as law and delivering something the dominant logic couldn’t even imagine.

The Disruption Canvas

There’s no shortage of canvas-style tools in the world. The Business Model Canvas (BMC) is the most widely adopted (and for good reason). If you’ve already defined your target market and monetization strategy, and you need to articulate how the parts of your business fit together to deliver value, the BMC is the right tool. It helps teams align on what exists and optimize the mechanics of delivery.

The Product Go-To-Market (GTM) Canvas is another example. It’s useful for teams who are ready to commercialize a solution and need to clarify customer journeys, value props, channels, and growth strategies. Other popular frameworks like the Lean Canvas and the Value Proposition Canvas serve critical roles in helping teams iterate on known ideas, refine target segments, and shape business models.

These tools focus on helping teams clarify and operationalize an idea — usually one that has already passed some threshold of validation or internal agreement. But when you're aiming to build something disruptive, existing tools fall short.

The Disruption Canvas is optimized for something else: identifying whether an idea has the potential to change the game. It’s designed to help teams to investigate market blind spots, confront outdated assumptions, and find leverage in simplicity. It's a tool to help you strategically and systematically create a product that couldn't have emerged from existing market thinking.

Disruption Canvas

One Last Thing

Disruption isn’t loud. It’s not obvious. It doesn’t announce itself with a pitch deck or a press release.

It starts small. It starts by serving people no one else sees.

Canva didn’t kill Adobe. They made Adobe’s map of the market obsolete.

Nintendo didn’t compete by making faster processors and games with better graphics. They opened the door to an entirely new audience of gamers.

That’s the playbook. That’s disruption.

I realize pricing is a boring subject. It shouldn’t be–hardly anything is more important for growth and leverage. But I’ll drop it for a while after this post… just need to cover the other end of the market before I get off my soap box.

If you haven’t signed up yet for our SF Tech Week event, we’ve got ~100 signups already. Sign up here. We won’t talk pricing–we’re talking about how to get your company acquired. And how to hang out with other founders/VCs and the S2S team. Would be great to see you IRL.

ICYMI Tech Week, (presented by a16z) is a week of events hosted by VCs and startups to bring together the tech ecosystem. SF Tech Week takes place October 6-12, followed by LA Tech Week October 13-19. View the full schedule at: tech-week.com/calendar

This post is brought to you by PicoMES. The next generation of manufacturers are building smarter, faster, and with fewer mistakes—and PicoMES is their partner on the shop floor. PicoMES helps innovative factories digitize and streamline assembly operations with modular software that scales as you grow. From digital worker guidance to error-proofed workflows with full traceability, companies as small as 6 operators to those as large as 1800 per shift trust PicoMES to bring startup speed and enterprise-grade reliability to modern manufacturing.

How to get your pricing model right in SMB markets

SMBs are a great place to start because for most startups, this is where you should start selling. SMBs move more quickly than enterprises and in some ways act more rationally, as we will discuss. There are, however, far more permutations of pricing models that apply to this market so it can be tricky to decide on one to begin with, and trickier to stick to it.

If you’re just picking this up, I highly recommend going back to the first post on pricing AI and SaaS products which covers some key concepts.

What Makes SMB Pricing Fun

You can’t build a company targeting SMBs without having a lot of customers. An enterprise SaaS company might be successful with 20–100 customers total. That simply won’t work in SMB land.

Lots of customers means you can experiment–something you cannot do safely with enterprises. Losing a deal doesn’t matter. There are a thousand of prospects just like it. That freedom makes SMB pricing the perfect laboratory.

You have to start your experimentation somewhere. Early on you’ll likely fall into one of two paths:

• Self-serve / low price ($20–$100/mo)

• Larger SMB deals ($1K–$10K+)

Either way, always err on the side of higher prices. We’ll get there. Let’s start with the first path.

Self-serve and PLG at low price points

At low price points, interviews and negotiation doesn’t matter nearly as much as experiments. You should do the homework we talked about in the first pricing post to start near the right price, but volume and experiments are how you build confidence in your pricing. If you’re selling something at $20–$50/month, you can test conversion sensitivity directly:

• Does $20 convert better than $25?

• Does $29 actually look cheaper than $30?

• Does $39 kill signups, while $35 is fine?

Stripe enables different payment links, and different products. For your first 10-20 customers you don’t have to worry about tying this all up with your back end (where it gets very costly and difficult to maintain). Instead do it manually, setting up different payments links and then enabling customers on the back end manually. Once either volume picks up or you have confidence, bake it into the product.

Here are some cheat codes based on a lot of lived experience (and consulting):

• There is usually not a lot of difference in conversion rate between $19 and $35. If a customer likes a product and it’s $25, they’re not going to chose the $19 competitor they like less just to save $6.

• Below about $199, SMB buyers behave like consumers and don’t agonize over price. They swipe the credit card and move on.

• There is a difference between $20 and $50, but not between $20 and $25 or $45 and $50.

• Past $199 willingness to pay starts to thin. People start to think. Expense limits kick in. Beware.

This psychology and behaivor is why most successful PLG companies underprice at the beginning and then expand revenue through multi-part tariffs: charging in more than one dimension. Dimensions include:

• Seats (new hires = automatic expansion).

• Usage credits (heavier use = higher bill).

• Feature tiers (unlock collaboration, security, analytics, etc.)

Take Bolt.new as an example: they charge for developer seats. If one developer uses the tool heavily, credits or seat upgrades kick in. Recently they added a third axis—team accounts. It complicates things, but it also creates more vectors for expansion.

At low price points, PLG + multi-part tariffs = growth. But don’t overcomplicate. Too many axes and customers will feel nickel-and-dimed. Base44 has gotten some pushback for having multiple types of credits. People don’t want to learn your complex pricing model, they just want to pay a fair price that is less than the value they are getting–and not think about it.

Higher-ticket SMB sales

If you’re selling in the $1K–$10K+ range, experiments alone won’t tell you much. Here, questions are your secret weapon:

• “What would you expect this to cost?”

• “How do you think about budget for solving this?”

If the prospect says $200 when you’re targeting $2,000, that’s not your customer. Be visibly surprised, talk it through–then disqualify. If they’re close, you decide: discount, move on, or hold firm.

This method works because, in SMB, pricing is rarely set in stone. A dentist with a 5-person practice and a boutique creative agency may both pay $5K/year for a scheduling tool—but for completely different reasons. You learn those reasons by asking.

Unlike the enterprise, you can absolutely close $10k/mo deals quickly in the SMB market. I bought data at my last startup–$10k/mo for 6 months–in a deal that took about 3 days to come together. Meanwhile I see companies trying to sell $5k pilots to eager, bought in enterprises–that are stuck in procurement and could take months to close. The high price SMB sale is truly the sweet spot for startup growth.

Beware the $200–$1,000 No-Man’s Land

Psychology matters. Buyers impulse-purchase up to ~$199. Apple knows this: their accessories hover just under the $200 mark. This is why most of your SaaS subscriptions–AI or otherwise–are also below $200 a month, at least on the sticker.

Meanwhile $1,000+ feels serious enough to justify an AE-driven sales process–and, the margins are there to support paying an AE $5-8k/mo if they can close and retain enough $12k/year customers.

The middle zone ($200–$1,000) is deadly. It’s too expensive for quick impulse buys and a true PLG motion, but it’s also too cheap for handholding.

If you find yourself in this range, you have three options:

1. Bundle more value and move up to $1,000+.

2. Unbundle to get an entry price under $200, with fast upsells.

3. Charge $1,000+ for the same thing. You’d be surprised how often this works.

The Bundling vs. Unbundling Dilemma

One of the hardest choices in SMB pricing is how much to bundle.Bundling reduces friction up front. One clear price, one package, easy to start. Unbundling reduces friction later. Low entry price, pay more as you grow.

At Skyp, we’ve learned this firsthand. Our most popular pricing ($1,250 as of this writing) is designed to cover most of what most customers want in one package. But customers always find edge cases—so we’re adding add-ons.

Avoid selling a car without a steering wheel. Your base package has to be usable and successful on its own, or churn will kill you. Though there are some high end markets where maybe customers will pay extra for a steering wheel (looking at you, Salesforce. And you, too, Porsche).

The key to what to include is success and the ah-ha moment. If keeping something out of the bundle means they will churn, then make sure it’s in. Put the customer first. But if there’s high willingness to pay for an add-on, keep it separate. And realize that when things are add-ons, competitive market pricing means less, because there is a trust and convenience factor involved.

Fyxer–an example of high prices winning

A great case study comes from Fyxer, recently profiled on Growth Unhinged.

Fyxer started as a services business, doing $5mm in ARR providing assistants to executives. They translated this knowledge into an AI-based product, but marketed it outside of tech to anyone who lives in email. They downplayed AI, and charged a huge premium over most AI-enabled email tools: $45/mo.

It’s a perfect example of why you should usually err high on pricing. Customers will pay more if the positioning and perceived value line up. Especially when you have a unique product, or are selling to a market that isn’t aware or familiar with alternatives. That higher price point has enabled significant performance marketing spend, which would not have been viable at a lower price point.

Onboarding and product are part of pricing

Pricing isn’t just numbers. It’s also friction. SMBs don’t have procurement departments and don’t want to spend months integrating your product. Instead, they want to:

1. Try it instantly.

2. See value in the first session.

3. Invite teammates quickly.

My PM friend Alex Leary calls this “push button, get bacon”. How long does it take between pushing the button–signing up–and getting the bacon–whatever you signed up for?

Bacon is different in each product. For Skyp, we decided that getting bacon means seeing the emails the AI generates. If you like them you’ll pay. If you don’t, you won’t. Worse–if it took you 20 minutes to get the bacon (like it does with apollo, lemlist, hubspot, etc.) then you lose interest. We’ve got it down to about 150 seconds. That winning get-bacon experience probably closes more customers than any discussion of the ROI of using the software.

Sometimes the person who wants to use your service isn’t the economic buyer. This is where onboarding is key and a free tier or free trial might be helpful. Stripe does this, enabling developers to just go rogue and build out payments with Stripe APIs. Later when the business folk get involved, they may want some other solution–but Stripe already won. Be careful here–but also be aware of your customers’ journey and who exactly cares about onboarding most.

Err on the side of high prices

The temptation, especially when you’re new, is to underprice. “If it’s cheap, more people will buy.” Maybe. But more often, low prices kill you.

If you’re competing with Intuit, and Quickbooks is $65/mo, do you really thing someone’s going to choose your solution because it’s $20? There are so many other reasons why they won’t pick your solution that are more important than price. If someone gets past all of that–you need to get paid! This is why Puzzle charges $50 or $100. Not $20.

Low prices kill you in so many ways:

• They make the product look less credible.

• They attract the wrong customer segment.

• They leave no margin for support, onboarding, expansion, or (later) CAC.

• They make it low-stakes for your customer, so they invest less time in getting it to work.

This last point bears repeating. If I’m an SMB and I spent $50 on your service, and it doesn’t work, fine. That’s your problem. But if I spent $5,000, goddamnit it’s gonna work. I’m betting my company, or my career, on that product working. This manifests itself in obvious ways. That $50 customer may not return–or even open–your emails. That $5k customer will pick up the phone every time you call her, will answer every email, and might even email you proactively so she can be more successful with your product. Not to mention–you can build a business with 20 $5k/mo customers. You are nowhere with 20 $50/mo customers.

You can even charge for things that you thought were free. One company I worked with had a motion supporting hardware companies bundling their software. Before the company brought me in, they did the work for free in a combination of software and services. It was killing the team, and taking engineers away from revenue-generating activities. We had 20+ clients asking all sorts of random stuff at all times of the day, needing software support, asking for more features. Constantly.

We created a program and charged $20k for the same service we were previously providing for free. Not only did we get rid of all the time wasters, but also we closed ~$100k in revenue in the first month. Don’t underprice your knowledge.

Speaking of knowledge, Elena Verna, now head of growth over at Lovable, wrote about pricing in her solopreneur journey. She started consulting for free, then charged $100 an hour. She moved through $200 and thought she had it figured out when someone said yes to $500/hr. Then she asked a friend what he thought she should charge. He said $4,000!

She tried it–and it worked. Same service. Literally 8x more money. What’s not to like? How would 8x more revenue change your business?

The Playbook for SMB Pricing

This was a long post. Because pricing is important. And SMB pricing is the core of most startup growth. Most companies will close more business with SMBs than with enterprises in the first 1-2 years. Getting pricing right can be the key to everything.

Here’s the tl;dr on how to get SMB pricing right:

1. Decide which path you’re on (self-serve PLG vs. higher-ticket SMB).

2. Experiment aggressively if you’re under $200/mo.

3. Ask the right questions if you’re over $1K.

4. Avoid $200–$1,000 a.k.a. swamp of startup death.

5. Use multi-part tariffs (seats, credits, features, bundles) for expansion.

6. Don’t sell without a steering wheel. Bundle enough for success.

7. Position premium when you can. Fyxer is the proof.

8. Keep onboarding effortless. Fast time-to-bacon.

9. Err on the side of pricing high. You can send me the extra $$ if you feel too guilty to keep it, but I think you’ll get comfortable with it.

Get these right, and SMB markets can be your growth engine. Get them wrong, and you’ll drown in churn and underpricing.

Want more on this? Go back to last week’s issue on enterprise pricing or the first in this series on pricing AI and SaaS products. Kyle Poyar’s Growth Unhinged hits on pricing a lot, as well–it’s important!

Next week I promise a change of subject. In the meantime, go get better margins!

Join us at SF Tech Week! Sign up here. Maybe also reply so I’m sure to approve you–lots of people are signing up.

Welcome to Seed to Sequoia – the newsletter for founders on GTM, strategy, and VC. We’re doing our second annual meet up at SF Tech Week–this one is a talk on how to get acquired. Sign up here.

This post is brought to you by PicoMES. The next generation of manufacturers are building smarter, faster, and with fewer mistakes—and PicoMES is their partner on the shop floor. PicoMES helps innovative factories digitize and streamline assembly operations with modular software that scales as you grow. From digital worker guidance to error-proofed workflows with full traceability, companies as small as 6 operators to those as large as 1800 per shift trust PicoMES to bring startup speed and enterprise-grade reliability to modern manufacturing.

Pricing startup AI and SaaS for enterprises

Enterprises are not just larger SMBs. Enterprise pricing is different and requires a different set of skills. If I was all about SEO optimization I’d write a list of the “72 ways enterprise software pricing is different” but I won’t waste your time. It’s really just these three:

1. Every enterprise customer is different. The decision process is different. It is rigid, unchangeable, and almost always long.

2. Enterprise buying motivations are very different than SMBs. Enterprises buy what they want when they want to, frequently acting irrationally–spending insane amounts or not buying something with obviously insane ROI for obscure, irrational reasons.

3. Experiments are impossible. Sales cycles are long so learning is slow. Deals are huge so blowing one is costly. N is very small and each company is very different, so even if you do experiment you’ll never have any accuracy or statistical significance.

As a result, your initial pricing strategy has to seem confident but be adaptable. You have to preserve high margins, because the enterprise motion is insanely expensive. You need to find and exploit the value metrics that matter the most to your customers, in a straightforward but enterprise-friendly way. Pilots are common, and pilot pricing is fraught with risks.

Why enterprise pricing matters

Before we get into the aspects of pricing for the enterprise, and how to come up with actual prices, it’s first important to recognize how important pricing is for startups selling to enterprise. Let’s take two examples of real YC companies (which will remain nameless).

One did not take pricing seriously, instead focusing on logos and quick wins. They ran $10–25k pilots, then scrambled for $50k subscription deals with some of the world’s biggest enterprises. Accounts did expand — some to $300–500k ARR — but many churned as the champions were either not bought in, or not in a powerful enough position to ensure success. Growth stalled, sales teams churned, and Series B metrics never materialized. It’s still a business, but a tired one.

The second went to the same market, with a different product and buyer persona, and set a $1M minimum. Anything less, and they walked. That company rocketed to dozens of seven-figure ARR customers, predictable growth, and optional profitability.

Different products make a perfect comparison impossible, but if I had to pick one factor that changed these two companies’ trajectories, it’s how they handled pricing for enterprises.

Every enterprise customer is different

Pricing for enterprise is challenging because each company is very different. McKesson, for example, allows director-level and above to sign off on $250k projects without getting anyone else’s permission. If you set your pricing for McKesson’s threshold, will it work for others? Maybe. Maybe not.

The buying process is also very different company to company. Much of it may depend on your price point, or your product. Some enterprises push down buying down to lower levels–or even outsource it entirely to agencies. Others move slowly and have multiple approvals required for even a small purchase.

The only way to win consistently is to get good at the art of enterprise selling. Understanding their business and the individual human motivations is key. You must learn who the decision makers are, what their motivations are, and what the process is for each customer. Never assume that because you sold to one company in an industry, you know how they all work–because you don’t.

If you get the selling process wrong, you may think you got pricing wrong. Or you may become insecure and charge too little. Recognize that each enterprise is different.

Enterprise buyers buy for different reasons

The common thinking is that enterprises are more rational than SMBs because they have systems in place–procurement, finance, etc. This overlooks one obvious fact: humans are making the buying decisions. Humans behave like humans, everywhere. They act in their self interest, make decisions based on emotions and later justify those decisions with rationalization. Humans are also fundamentally social creatures–which is another way of saying that company politics matters.

You have to really understand your buyers to understand why they will buy. Some are experimenting. Maybe they work for the Strategy department or some other non-P&L owning group. These are the worst possible customers.

Others are buying because their career is on the line. If they don’t have an answer for X problem, or Y corporate directive, they might lose their job or miss out on a promotion. Understand the very human emotions behind your buyers, where they fit in the organization, in order to set prices. $20k may seem a lot to you, but if the GM of a $1bn division is betting his or her career on it, he might prefer to pay $2 million–for the same product.

Do not experiment on enterprises

Experimenting with enterprises is impossible. There are too few that are similar to ever have statistical meaning. Usually the value of winning any one enterprise is so high, it’s not worth risking losing the deal in the name of learning. And the time it takes to run a process–which might be a year or more–means that if you find out something doesn’t work, you’ve lost a year. Which for a startup might mean you are now out of runway.

Instead, focus on winning the sale at a high price. It’s tempting to discount for your first customer or two, but avoid this. If your product provides value enterprises will be willing to pay for that value, and if they are not willing to pay then they are wasting your time. If you get a high price from one, that might be all you need to build your company. Plenty of successful companies have started with just one major enterprise customer and reached profitability on that one contract.

How to set prices for early enterprise sales

There are two key steps to pricing for enterprise sales. First you have to understand the customers’ business, and how your product fits into that business. Don’t do what I did and waste 5 years trying to sell something nobody wanted at a price nobody could afford.

You need to understand the business impact your product will have and where it fits from your customers’ perspective. To do this you really must understand the business. Waving your hands and saying you’ll “decrease costs 50%” or “increase leads 20%” or whatever works on angel investors at YC demo day, but it never works on actual enterprise customers.

Here are some good questions to ask that validate the problem space and help you understand how your customer views the problem and value:

• What’s the trigger that makes this problem show up for you?

• Who feels the pain the most in your org?

• How are you handling it today?

• Where does it usually break down?

• What happens if it doesn’t get solved?

• What outcomes matter most if this goes right?

• Which metrics or KPIs does leadership actually track here?

• What do you spend today (time, people, tools) to handle it?

• Where would budget for solving this come from?

• What would make this urgent to solve in the next quarter?

You may notice a light “jobs to be done” theme in those questions. It is easy to get absorbed in big picture value, but when you are figuring out both what to build and what to charge for it, understanding where it fits in the process is crucial. Knowing this will also help you sell, as your questions will be informed and reflect a deeper understanding of your customers’ perspective. You won’t have to prove you can do it–they’ll assume you can based on how good your questions are.

Once you understand the customers’ business you should have a sense for the value your product creates. You can then start having sales meetings, to both validate and test this value. You are not experimenting in the SMB or consumer sense. Think of it more as a negotiation.

Here are some good questions to ask that help the prospect articulate the value themselves, in their own words:

• How are you solving this today?

• What happens if you don’t solve it?

• Where does this sit on your priority list?

• Who feels the pain the most internally?

• What outcomes would make this a no-brainer?

• How would you measure success with us?

• What budget do you typically allocate to problems like this?

• What’s the cost of doing nothing here?

• If we achieved [the goal they stated], what would that be worth to you?

Build your enterprise pricing structure

The key to your pricing structure is the value lever. It can be simple at first. You don’t have to have a 3 part tariff out of the gates. You do have to capture significant value. Getting the value metric right is the most important thing at this stage.

For example if you work with law firms and create value for each case, perhaps charge per case. If each case they work on generates an average of $2 million in fees, you should be demanding tens-of-thousands per case. Not $200/seat. If you throw out $50k per case and they come back with “That’s ridiculous, how about $20k” you have won. On the other hand if you say how about $50k per case and they say “you’re nuts, we never pay per case” then you did not ask enough questions earlier and you have failed to understand their business. Of course, if you throw out $200 per seat and they jump on it–that’s a bigger, more costly failure.

If this is an established area with established vendors, usually conforming to existing pricing is the best path. It is faster than educating the market on new pricing, companies may already have allocated budget to it, and these sorts of purchases get less scrutiny because they may be viewed as just a cost of doing business. For example if you were selling legal services as an AI-enabled law firm, and billed by the hour, you might have double the margins of non-AI law firms but buyers could easily buy this. Instead of trying to convince an enterprise to pay on some totally different scale (even if that new scale makes your product or service cheaper).

Make sure the prospect articulates the value math. You won’t win by presenting your ROI spreadsheet—they need to own the numbers. If you did your homework, you will have a pricing structure that fits with how they see the value of your product. You may capture quite a bit of that value, but it has to fit in that paradigm. Remember the rule of thumb: if you create $10M of value, you can capture $1M. AI has shifted this math slightly because it takes on even more of the work and can be more accountable for a successful outcome. If you can point closely to outcomes you might be able to capture 15%-25% of the value–but far from all of it.

Be comfortable with whatever motivations your pricing creates. If pricing per case means a law firm will only use your product on higher value cases–that might be better than pricing per seat, and using your product on cases that are low value where it is perhaps not worthwhile to use anyway. You can address ‘share of wallet’ and bundling later, once you better understand usage patterns.

Mistakes in early enterprise pricing

Given the low volume and high touch nature of enterprise sales, pitfalls are everywhere. Here’s how to avoid the most common mistakes.

• Don’t underprice enterprise. Make sure you understand how the buyer buys, what similar things they pay for, and the value you create. You don’t have to price on an outcome basis to capture value, but realize that low prices can be a negative signal to enterprises.

• Beware cost-savings pitches. Enterprises don’t fire 500 people and reallocate budgets to your AI tool. Sell transformation, revenue growth, or avoided hires. At Skyp.ai, we sell SDR efficiency, not headcount reduction — and we never market “cost savings” because it doesn’t sell.

• Enterprise sales is expensive. Sales cycles take quarters, not weeks. You’ll burn cash on travel, events, and development even before you hire an AE. When you do, each AE may cost $500k–$1M including equity. If the pot of gold isn’t big enough, bad things happen: you have negative margins per customer, and/or VCs won’t fund you.

• Validate with the right buyer. Junior champions may love you but can’t buy. They may suggest discounting or lower prices because they lack perspective or budget. Push for intros to the actual budget owner. See also: What if your price is right but your customer is wrong.

• Avoid “strategy” titles. They may have a $50k test budget but zero ownership of P&L. They waste time with pilots nobody cares about. Real buyers own revenue lines (GMs, VPs, sometimes Directors at bigcos like Cisco).

• Don’t over-accommodate. Salesforce doesn’t change button colors for free — neither should you. Everything can be on the roadmap, but only paid customers accelerate features. No spec builds for prospects.

• Lost deals aren’t really lost. If a prospect only bites with free custom work, free pilots, or an experimental strategy team’s budget, you weren’t going to win anyway. Better to walk and focus resources on winnable, high-value accounts.

Pricing pilots

The purpose of an enterprise pilot is only to confirm whether your product meets the prospect’s needs — not to prove value you’ll have to renegotiate later. Always align on pricing for the full rollout before starting. I learned this the hard way: we ran a $25k pilot for Couche Tard (Circle K), only to hear afterward that they agreed simply because “it was only $25k.” We hadn’t had the deep conversation about ongoing pricing up front — a clear red flag.

The actual pilot price matters less than ensuring it covers your costs and signals seriousness. Enterprises routinely pay McKinsey $500k–$1M for boilerplate slides; there’s no reason to subsidize your own work. Charge a services or implementation fee if you can, and structure the pilot so it naturally converts — e.g. five cases at $20k each, then $50k per case unless they commit annually.

Anyone who can pay millions a year can afford something for a pilot, and if they can’t, that’s a signal. Usually you’re talking to the wrong person. Sometimes the perceived value isn’t there, perhaps because it simply isn’t or perhaps because of other, higher priorities. Ask questions, anchor on value and fairness, and remember: enterprises buy what they want, when they want. Your job is to make sure you capture full value when they do.

Pricing for enterprises early in your journey

Enterprise pricing isn’t an A/B test — it’s a negotiation grounded in how the customer actually creates value. Do the work to understand the business from your customer’s perspective, and what impact your product will have. Anchor on a value metric the buyer recognizes, and sell to P&L owners who can say “yes.” Get agreement on the price of the full rollout before you start a pilot, charge enough for the pilot to cover real costs (ideally with a bit of margin), and structure it so conversion is the natural next step. Don’t pitch cost savings. Instead, sell outcomes, avoided hires, and revenue.

Protect your margins and your time. Say no to free custom work, “strategy” or “innovation” budgets, and eager champions without authority. Walk from the wrong deals so you can win the right ones at the right price. It only takes one great enterprise customer to change your trajectory. You might be surprised to see walking results in a higher close rate–even on the deals you walked away from.

If this resonated, come say hi at SF Tech Week! If you want to hear about something specific (on this topic or others) just reply and let me know!

Welcome to Seed to Sequoia, the newsletter for founders on GTM, strategy and building companies. We’re excited to host an event at SF Tech Week–How to Get your Company Acquired. We’ve partnered with IPG and Skyp to host it in the heart of SoMa. Space is very limited, so if you’re interested please sign up ASAP! Hope to see many of you there!

And now, onto one of my favorite topics: early stage pricing.

It’s easier than ever to build and launch a software product today. The tricky parts remain figuring out how to get people to use it, and what to charge for it. Pricing is particularly fraught now that it seems like two smart people armed with Cursor can build almost anything.

How AI has changed SaaS pricing

AI-enabled products can deliver a lot more value than SaaS without AI. Therefore it can capture more value. How much more? If you look at the relationship between SaaS and one-off license-based software, SaaS captures about 1.5-2x more value. So it makes sense that AI enabled products might capture 1.5-2x more value than their SaaS equivalents.

Looked at another way, if in the old days, using software licensed from Oracle and set up yourself creates $100 million in value, Oracle might capture maybe $5 million of that value. Because the customer has to do a lot to get that value–buy servers, set up a datacenter or rent colo space, and probably integrate that with their other systems. They might even pay $5 million for a few years before ever seeing any value from the software.

SaaS changed that. Salesforce, for example, made it so the customer didn’t have to do most of that work. Just configure and get value (in theory). As a result the equivalent SaaS might get $10 million of that $100 million in value, 2x more than the legacy Oracle. Humans are still doing the work–sending the emails, taking the sales calls, updating the CRM, etc. But it’s a lot faster to value and the vendor does a lot more while the buyer does a lot less.

AI takes that one step further, because it can accomplish some of those tasks. It can take notes and update the CRM. It can write the emails (see also, Skyp). It can even pick up the phone. The customer does less, the AI does more. As a result, if its use generates $100 million in value for the customer an AI tool might capture $20 million of that value.

But how much value? How do you structure it? Those problems haven’t changed. If anything, they’re more difficult as the landscape is changing so quickly and it can be hard to explain why a customer should pay $20 million to replace a tool that they had been paying $10 million to use.

Here are some tactics and structure for figuring out pricing. It turns out this is a big topic, so unless I get a lot of emails that ask to not hear more about pricing there will be multiple posts in this series. This first post covers the principles and tactics to setting prices for all markets.

Take PMF for granted

Well–don’t, actually, but for this exercise let’s assume you have built something people want. This is a major assumption. Do the work to make sure someone actually wants what you’re building. You’re ready for this when a prospect who sees your solution (in a demo, trial or pilot) tells you that it will solve their problem, and asks what it costs or asks how or when they can start using it.

Pricing Principles

Pricing can feel overwhelming. You can literally charge anything. Palantir has billion-dollar contracts. The App Store supports prices down to $0.99. How do you decide where to be on an infinite continuum?

In setting prices, there are three key principles that can make it easier to assess both strategies and actual prices.

1. Value Alignment — Price is anchored by delivered value.

2. Fairness — Customers need to feel the price is justified.

3. Simplicity — Use the simplest model that maximizes your goals.

What matters most is the perception of your customer–at all levels, from B2C to large enterprises. If that $0.99 app seems like a good value, great. People will buy it. Same with the billion dollar team of forward-deployed AI engineers from Palantir. If the customer perceives they are getting a value from what they pay you, they will buy. It may be more than they want to pay, but it is (by definition) something they can pay.

They must also consider the pricing “fair” – which is less rational. Fairness exists both within the scope of your own pricing, and across the market and competition. How many banks advertise “Free Checking”? It seems unfair to customers charge people to maintain a checking account. The fairness principle does not have to be rational; the only thing that truly matters is the perception by the prospect or customer.

Simple pricing usually wins. Nobody wants to learn your fancy pricing model–not even enterprise procurement teams (even though this is, basically, their job). Complexity makes it harder for you to implement, maintain and experiment. Complexity adds barriers to new customers feeling comfortable buying and extends sales cycles, resulting in slower growth. It also raises the likelihood of a customer gaming your pricing in a way you did not foresee. And nobody likes surprises. While you might think it would be great to send a customer a surprise $1 million overage bill, this is not a winning strategy. Because, among other things, it is not considered fair–after all, you invented your complex pricing scheme that led to this overage; they did not want to bother learning it.

Implementing pricing principles with tariffs

Pricing nerds call the way you extract value a “tariff”. There are one-, two-, and three-part tariffs that encompass pretty much every possible way you can charge for SaaS and AI software:

• One-part tariff: Single fixed price (e.g., flat monthly subscription, $99/mo).

• Two-part tariff: Fixed fee + variable usage (e.g., $50 platform fee + $0.05 per API call).

• Three-part tariff: Fixed fee + allowance + overage (e.g., $20/mo includes 10 million credits, then $30 will buy $10 million extra credits–this is Botl’s pricing today).

The more complex the tariff the better it can align value, and possibly appear to be fair, but (obviously) the more complex to administer–and explain. That said you can absolutely have a simple, easy to explain three part tariff. You probably are paying for many services that do this–OpenAI, Claude, Cursor, Bolt and Lovable all use this model.

While these concepts have been around for a long time, the rise of AI applications with significant marginal cost (unlike SaaS of, say, 2 years ago) has led to a lot more use of more complex strategies.

Fairness is critical. Pricing must be fair across your pricing and customers, and in the market. For example you cannot charge Customer A $10,000 and Customer B $100,000 for the same product. While you could maybe do this, it will cause havoc within your company. Salespeople will be confused. Marketers will be confused. And if your customers ever learn what happened, guess which one will be angry? Hint: Not the one paying $10k. Similarly if you charge $10,000 for a feature that a competitor charges $100 for within your market segment, it will cause problems.

An example of this is Rippling and Deel’s pricing. Rippling charges customers a fee per seat per month–even if there is nobody in that seat. Deel, which started as a way to manage offshore freelancers, only charges in months when a payment is made to an employee or contractor. That seems more fair. If I am not paying the contractor why do I have to pay a fee the payment software? Even when Rippling is actually cheaper, it can feel more expensive because of this fairness principle. Obviously Rippling’s platform is available to that contractor even when they are not being paid (end of year taxes, etc.)–but if the customer feels it is unfair, it is unfair.

Two steps to setting prices

First, you need to understand your customers’ business. For example, at Skyp.ai I know startup GTM extremely well. I know what companies pay SDRs, AEs, marketers. I know all of the other costs–the recruiting, management, tools, etc.–and I know the emotional cost of bad hires. I also know the value of getting leads, and how companies think about and value them, and what alternatives exist to using Skyp.

If you don’t know your customers’ business that well–you need to get there! Talk to consultants, talk to customers, do whatever it takes to understand both the financial and emotional aspects of how that customer’s business works. Advisors or advisory boards can be great resources here.

Then, once you understand your customers’ business, you need to ask questions and start getting paying users before you create a repeatable pricing model. Expect to be wrong, so don’t overthink it. There are two ways to do this. Both can be done with intention.

Ask the three primary pricing questions

One strategy is to ask the three main pricing questions in a survey or interview format.

• “What do you expect this to cost?”

• “At what price would this feel too expensive?”

• “At what price would this feel too cheap to be credible?”

This will get you the best data, especially in an open-ended interview format where you can ask follow ups. You’d be surprised at how high the “too cheap” price frequently is, especially for large enterprises. A good reminder not to project your values on your customers. Usually you end up somewhere around the “too expensive” price.

The limitation is that these questions feel out of place in a sales conversation because your prospect will think (correctly) that you definitely have no idea what to charge. This strategy works best in a product development or discovery phase.

Ask three pricing questions to close early customers

The other strategy focuses more on closing sales (i.e getting real data) than the more scientific goal of pricing research. If you are in prospective sales discussions, you want to ask questions that are more focused on value, precedent, and their business.

• “What do you expect this to cost?” 

• “How do you currently budget for (or buy) tools/services like this?” 

• “If we were discussing renewal in 12 months, what outcomes would you need to see to justify the price?” 

The first question frequently leads to unexpected answers. For example with Skyp.ai I heard answers of both $29/mo and $10k/mo in the same day. It just depended on the lead and where they were coming from. Scrappy founders using Apollo lumped Skyp in with cheap GTM software. PE backed enterprises and late stage startups paying 40 SDRs saw it as a way to save on six-figure monthly payrolls. Obviously this has implications for who you target. Be sure you are choosing the right customers.

Asking about budgeting serves two purposes. First, it shows you what tools or services your customer views as analogous. This is far better than what ChatGPT told you–it’s the customers’ own words. So if they tell you this is like buying Salesforce which cost $1 million, then you can reasonably understand they think what you’re selling is going to cost $1 million. It also tells you about how they actually buy things like this. For example, maybe it took 9 months to buy Salesforce, and required allocating for it in the annual budgeting process. Follow up questions here are key.

The renewal or outcome question anchors your customer to the value you create. If you create that much value, then they can reasonably agree to pay you for it. For some companies, the ability to not renew is enough safety. This question creates context for you and anchors the prospect to the value. It also is great to revisit throughout the period to make sure you are delivering on the metrics they care about, so when it is time to have that renewal conversation you can check this box.

Ask questions that anchor to higher prices

You can ask additional anchoring questions, if you have some idea of what you want to charge. For example, “What if I told you this cost $1000 a month?” can lead to an interesting conversation. Really dig. Validate your understanding. If you are in a sales process you care more about getting a customer to pay the price you want. Not only is it revenue, it is a much more valuable learning than mere words. Anchoring and tying the price to the value delivered is more important than open-ended questions.

A favorite tactic is to frame around the fear of loss: “One of my founder clients paid an SDR who didn’t work out $110k a year. Is that what you pay your SDRs?” When you follow up with a $1k/mo price, it seems reasonable relative to the 100x larger number.

Never underestimate the power of anchoring. If you lead the call with “Look this isn’t crazy expensive–not like $20k a month or something–but I want to make sure you’re not expecting a $2k/mo cheap tool,” nobody will tell you they think it’s only going to cost $29/mo. Some might hang up – but that saves you time. Subconsciously, as you go through the demo they’ll look for justification that it’s a Range Rover and not a used Hyundai. They’ll see what you told them to see. Your vibecoded React front end will feel expensive, not cheap.

Maybe wing it–but not for long

For your first ~3-10 customers, it’s fine to just wing it. You’re going to be wrong a lot because you simply don’t know enough. Having customers is far better than not having customers, and when winging it almost every founder under prices. So you’ll be fine.

For the first customers, the only data point that truly matters is: did someone actually pay this price? Be wary of relying too much on answers to questions if nobody is buying. Price is very much part of PMF; you can easily have positive demos or trials, and not sign customers because of your price or pricing structure.

As soon as you start see a pattern it’s time to build a pricing model. You do not have to implement it yet in code–that can come later. Setting up your pricing model unlocks a repeatable sales process and gives you the ability to experiment effectively.

If you have a dozen or more customers and you are still winging it, you are undoubtedly leaving money on the table. You need to be sure you understand your customers’ business, the value you provide (from their perspective) and charge accordingly. The earlier you do this, the more success you will experience–not only because you did your homework, but because you’ll bring the confidence of being well prepared to all of your sales meetings. And that is priceless.

Hope you enjoyed this newsletter! Next week we’ll have more on pricing. Interested in building a company that gets acquired? Sign up for our event at SF Tech Week. Want tools to grow your business now? Check out Skyp.ai, AI microcampaigns that deliver results–for founders and early GTM teams.

And of course if you want to just talk pricing–reply! Love hearing from you.

Welcome to Seed to Sequoia–the weekly newsletter for founders. Today we’re hitting on M&A, always a favorite, but from the opposite angle: how to buy other companies. There are deals out there, and if you’re running a tight ship or sitting on a war chest you might consider taking advantage.

On the other side of that coin–I’m speaking at a TechWeek event on Oct 8 in SF: How to Get Your Company Acquired. It will fill up fast (we haven’t announced to the public yet) so if you’re inclined–sign up early, and feel free to share the invite with founder friends and portfolio companies. We may have room for 1 more cohost–if you’re interested, just reply to this email.

Mergers and Acquisitions (M&A) is always a favorite topic of founders. Some VCs will tell you that it’s IPO-or-bust, and I respect that approach. If you’re raising VC funding that is usually the game you are playing. But great outcomes come through M&A, probably the majority. And no companies are doing an IPO in their first 5 years, so those exits are exclusively M&A. E.g., Base44–six months to acquisition. It takes 6 months to think about preparing for an IPO, let alone prepare.

If you want to nerd out on stats, be my guest, but the numbers are ambiguous and somewhat irrelevant. Carta reported 642 M&A transactions in 2024, based on its data (which only includes companies on Carta). While in the same year there were around 1,133 IPOs globally(according to World Federation of Exchanges), many if not most of those were not tech startups. For example, the parent companies of Viking Cruises and Salomon and Arc’teryx were among the 2024 class of IPOs.

Looked at another way, when I was at Opendoor we did 4 M&A transactions and had done two previously. That is 1 company that IPO’d, but 6 M&A transactions. Does that ratio hold across all companies? No. It’s probably higher, in favor of M&A.

The M&A market changes over time. When I ran M&A at Opendoor in 2021, it was a seller’s market. Lots of M&A was happening and it became contagious. With zero interest rates, venture valuations were at all-time highs, driving M&A activity (waiting was expensive!). Opendoor used M&A to solve for how hard it was to hire great engineering talent; others were solving for other challenges. Boards liked dealmaking and suggested it to their founders and leadership. They were much quicker to approve deals, sometimes at absurd prices.

What about today?

My, things have changed

Things changed for the worse in 2023 and they stayed changed, especially in the market for sub scale or failed startups. In 2024 I helped a good friend try to sell his YC startup. He had amazing tech, a couple of customers, and had run lean for a long time. It seemed like they had found PMF, or at least were close enough it was worth an acquihire to the right strategic.

I introduced him to the CEO of the perfect acquirer. Large, well capitalized, fast moving. But their CEO couldn’t take it to his board. Rounds-to-zero revenue was a nonstarter in 2023. Even though he’d be taking on just 2 employees (the founders), it wasn’t worth it. Even though the risk was near-zero, it still couldn’t be justified. Sucked for everyone.

The biggest change is that fundraising for non-AI companies has become very challenging. While venture investment numbers are staggering–they are almost entirely going to AI startups. In the land of regular VC backed businesses, where hitting $10mm ARR over the course of years–not weeks–is a big deal, fundraising remains very challenging to impossible. The biggest, best VCs only want the $10mm-ARR-in-weeks investments. And smart founders don’t want to bring on low-tier or “venture tourists” because they do stuff like this.

Growth expectations have risen, while in some industries headwinds have become substantial. Take climate, where Biden administration largesse has turned into arbitrary shut downs of already permitted and half built projects.

So what can you do?

M&A as a growth engine

M&A can be your growth engine. Inorganic growth is another way of saying growth via acquisition, and it works. Let’s say your company is doing $5mm ARR. Another company in your space sells to a slightly different but similar ICP and also has around $5mm ARR. Perhaps they’re downmarket; perhaps they’re upmarket; perhaps its an adjacent market.

If you merged, a few things change:

1. You are now big enough for larger funds to care. There is more than a 2x difference between $5mm and $10mm in ARR.

2. You are now big enough for financial acquirers (PE) to care. The juice is not worth the squeeze under $10mm ARR.

3. You may find a better talent pool to unlock more growth, including (possibly) a replacement CEO or other c-suite level person.

4. You may be able to find some economies of scale. You probably don’t need 2 CROs, 2 CMOs, etc.

5. You doubled your revenue, which always looks better than flat growth in a pitch deck.

There are advantages to this strategy–and disadvantages. But it is different than selling to a corporate acquirer in an actual exit, and it is different than a corporate acquirer buying your company.

True mergers

While big companies buy small companies for the new hotness, or growth, or because it’s cool–startups merge because the combination is better than separate, competing entities. Sometimes this is for “economies of scale”.

One of the best examples of this is oDesk–eLance. They were arch rivals whose CEOs regularly got into public fights (before that became commonplace). But with nearly identical solutions, they realized that merging the platforms could help them grow more quickly and expand the market. They merged in 2013-2014 and in 2018, went public as UpWork.

Smaller companies can integrate more quickly and generally experience more economies of scale. This is because there are fewer management layers to begin with, and the overlap is obvious. Two CEOs. Two CMOs. Two Salesforce instances. When you are losing money every month there is every incentive to cut costs quickly. The trick is doing the integration quickly, and not let it distract you (which it did at UpWork, to an extent).

If done right, the merged entity comes out of the transaction leaner and stronger. The best people in each position have been kept, the weaker let go. Efficiencies have been quickly found with fresh eyes and acted upon. M&A can also be great for marketing–creating more market awareness and excitement.

Tuck in acquisitions

The other way to grow inorganically is to buy smaller companies or teams. This gets tricky. Lots of smaller startups who are realizing they are off-track for Series A are also burning a fantastic amount of money to try to make their Series A metrics. These are not viable acquisitions.

Worse, many would be acquirers in the $5-15mm ARR category are also losing money every month. So adding a money-losing business to a money-losing business simply means losing money faster. You might convince investors (new or existing) to back such an enterprise, but you will have to really tell a great story.

Instead, consider finding a team of 2-5 hungry founders who are tackling an adjacent problem without the sales or marketing acumen (or luck) to get traction in the market. Maybe you can bring that hunger and know-how onto your team for not very much (if anything) up front. You might even add a couple of customer logos. This is where a lot of the opportunity is in today’s market.

How to set yourself up as a target

If you’ve realized you’re at the end of your runway, first go read this post. Your best option is selling to a cash-rich acquirer. It is not necessarily better to be acquired than it is to simply shut down. Here’s more on the shut down process, if you’re in that spot. A merger or equity acquisition can be a solid option, and better than shutting down, in many circumstances. I maintain that the rise of AI coding tools like Cursor also makes it viable to sell your IP without your team to run it, which opens up new possibilities.

Build relationships with partners and potential acquirers from early on, so that if you are in this difficult spot you have people to call who can start a process from a position of trust. Then, make sure you have runway. Be realistic and unafraid to cut costs. It’s very hard to dictate the timing of a transaction, and anyway nobody wants to buy a liability–and if you have a high burn, your company is effectively a liability in the eyes of your acquirer’s finance department.

How to run M&A as an acquirer

If you’re a startup looking to acquire companies, first off recognize that every single acqusition is a “bet the company” activity unless it is truly an acquihire. M&A is not something to be taken lightly or to learn on the job. Why?

1. It will distract your management team completely for at least 2-3 months.

2. It will involve significant cost (cash or dilution or both)

3. It will change the culture and composition of your team

4. It may bring on company-ending liabilities

5. Your board will judge you for it, as boards judge every CEO doing M&A

All of that considered, acquisitions can be a great idea. Many acquisitions made companies. Zynga was famous for acquiring popular Facebook apps with all of the cash it minted from its early success. Opendoor’s acquisition of Pro.com helped it reduce costs in its entire business and weather the downturn in the real estate market.

Here’s some quick advice on how to set yourself up for success in M&A.

1. Get board buy in first. Do not go rogue. Best case it slows down deals. Worst case you pour a lot of time into a deal and then your board shuts it down AND fires you. (I did say “worst case”)

2. Get good advice. Interviewing bankers or advisors is free and will teach you a lot. If your CFO or someone on your team is experienced, great. Fractional is also great–I do this work and many others do, as well. Get intros from your board and experienced founders you trust. Do not learn this activity on the job. There’s a reason why advice is expensive: a small mistake can literally cost millions.

3. Run an actual process. Do not just buy the first company you come across. Look across the entire market, possibly globally, and talk to multiple companies at once. Some will have ridiculous expectations. Some will sell for peanuts. You may have choices–you may not. Only one way to find out.

4. Once you find a target, do real diligence. Like a customer at Bloomingdales–you’re the buyer. Make demands. Get the data you need, in the format you want. If you get weird vibes–walk away. If you were looking for XS and this is an S, walk away. Your advisor should handle this for you, with help from other professionals (lawyers, accountants, etc.). People do diligence even in a “sellers market” (which this is not) so don’t feel bad about asking tough questions. Ignorance is not bliss.

5. Plan the integration. This is where even experienced, public acquirers fail all the time. Make sure you as a CEO and company focus on getting the integration right–and do the work. There is no magic-hand-waving that makes this easy. It’s tough and a LOT of work. Far more, usually, than getting a deal done–and for longer.

The word that sums all of this up best is “discipline”. It can be a long process and deal fatigue is real. But you need to keep disciplined the whole time. If you find some show-stopping finance issue on day 54 of due diligence, stay disciplined. Even if you thought you were 48 hours from closing, you may need to call the deal off.

If you find the dream startup with the perfect product but $1mm net burn a month–and a CEO who wants a 15x revenue multiple–maybe lose a night of sleep, but pass. Stay disciplined.

The M&A window may be open

Back in late 2024 I had made my peace with the election results reasoning that at least the casino will be open. Turns out I was wrong. But the M&A window may be opening again, in a different way. Boards are looking to land their underperforming companies, and extend the growth of the ones that are doing ok if not great.

Climate and clean tech are particularly affected. The negative perception has probably exceeded reality. EV charging stocks are all in the dumps–Chargepoint down 99% from its peak–even though there are more EVs on the road than ever before. If you’re disciplined, there are deals to be had. It might be bet-the-company, but if it’s a bet you can win… it may be worth taking.

Welcome to Seed to Sequoia. We’re hosting an SF Tech Week event–How to Get Your Startup Acquired. Would love to see you there!

I spend my time talking to founders, early stage investors, and general startup junkies. One of those friends–probably the most helpful angel investor I know, Huan Le–mentioned a question he asks founders the other night.

I think it’s the best question ever.

The best question ever

“What company or person do you want to be your customer?”

It pains me to admit, but this is even better than “Who asked for this?” Which is also a great question, and one of my most popular articles.

Why is this such a useful question?

The investor’s perspective

Huan is an incredibly helpful investor. He’ll roll up his sleeves and do sales calls. He’ll hit up the ex-Bain mafia and get meetings with pretty much anyone for one of his companies. He’ll bring in other investors and help you raise your next round.

But the biggest help, by far, is helping you get customers. How would he know if or how he can help if he doesn’t ask who you want as a customer?

If you like magic wands, you could frame it that way: “If you could wave a magic wand and get any customer you wanted, who would that be?”

The question comes from a place of helpfulness. If Bob at Walmart is your perfect customer, and I worked at Bain with him, I can help you. If you want a hospital COO as a customer–maybe I can’t help you, because I don’t know any of them–and I’ll let you know.

It’s a great way to understand the product and who it is for, or where the founder wants it to go. It can also help identify the gap. If you pitched an AI chatbot for small ecommerce companies, and then answered that question, "Walmart”–we have a problem. Maybe you don’t get Huan on your cap table.

The operator’s perspective

You need to be able to answer this question clearly and succinctly to run your company. There’s a lot of talk about ICP (ideal customer profile). It’s everywhere. It’s in this newsletter a lot. Because it should be! Figuring out your ICP, finding them, engaging them, building a painkiller they need–that’s the entire game. If you don’t know exactly who that is, everything will be a struggle.

The thinking on ICP gets complex. People talk about personas. They do customer journey mapping. They spout jargon. But the simple question is, “who do you want as your customer?”

From that perspective, a lot of things become clear. If that ideal customer is Walmart–spending time building Shopify plugins is probably not time well spent. If that customer is a Shopify seller, spending time in Bentonville, AR is not time well spent.

The team’s perspective

My friend Tod Sacerdoti turned his company Pipedream into a vibecoding powerhouse (and rebranded–string.com). They’re getting a thousand signups a day. I’m sure Tod can tell you who the perfect customer would be, who he’d wave his magic wand and get. Just because you’re getting 1,000 signups a day doesn’t mean your perfect customer is signing up. So there’s work to do.

Being able to answer this question for your team–especially new team members–helps implicitly and explicitly. New team members, especially GTM team, join brimming with enthusiasm. They see lookalikes everywhere. If your product is great for rocketship makers, maybe it’s also great for satellite makers. Or plane makers. But maybe–just maybe–it’s only great for making rockets? If you can be clear about who your perfect customer is, your team will be able to channel that enthusiasm usefully.

Clarity

At TripIt I helped launch the API. Public APIs were rare. Twitter’s was the only well known one at the time; Salesforce’s was in the works, and about six months after this launch, Mikkel Svane asked me to discuss whether Zendesk should build one (I said yes–but I should have said “I’ll come do it for you”).

Launching one was a new thing. We sat down and thought about who we wanted as a customer. We went to the App Store, looked for the #1 app in Travel, and decided we wanted the press release to mention that they had integrated TripIt.

It was an app called FlightTrack, and so I called them up. Ok, I didn’t call–I emailed. When Ben, the founder, sent the NDA back I realized he didn’t have an entity, even though he was the top app at $4.99. So I sent him off to incorporate, and about six weeks later, the integration launched.

Here’s the TechCrunch story on it.

The fact that there is a TechCrunch story shows how well this worked. That partnership, by the way, went on to be responsible for over half of our press mentions.

Your ICP may not be your customers

When 11x raised money they had a lot of churn (now very publicly in TechCrunch). But they knew they weren’t getting the “right” customers and focused on shifting their acquisition and qualification strategies to get better customers.

The startup GTM dream is organic traffic. That Reddit thread or ProductHunt launch that sends a firehose of thousands of new users to your app. Just like the TechCrunch firehose of 10 years ago, the vast majority (98%+) won’t be your ideal customer. And the sheer volume will make it impossible for you to find that 2% who are your ICP.

It’s helpful if you can at least articulate an answer to the question. It’ll up your chances of finding the needle (your ICP) in the pile of needles (the thousands of other customers who signed up and look pretty exciting).

Have an answer

As I write this I realize that for Skyp.ai I’m not sure I have a great answer. We have a lot of founders, and small teams. We have some enterprise customers, and more in the pipeline. We have some fractional people, using our Pro version to send microcampaigns. Many are getting great results. Some are not, but there’s no obvious correlation or causation.

But who is the perfect customer? If I got out my magic wand, who would I want as a customer?

That is the question. The best question.

Forwarded this email? Subscribe to get it delivered weekly 👇️ like 4k+ other awesome founders and VCs!

Welcome to Seed to Sequoia! I’m working on a post on LinkedIn metrics–if you spend time on LinkedIn please reply with (a) metrics you track and ballpark of what they are, (b) how much time you put into LinkedIn, and (c) what has changed lately (if anything). Got some amazing replies so far–looking forward to sharing the knowledge with everyone!

Oh, and if you were forwarded this, don’t forget to subscribe 👇️ 

The origin story as critical branding

The origin story is a powerful tool in the startup marketer’s arsenal. Startups are fighting an unfair fight, and need to leverage any tool they can. Many–perhaps most–miss a key opportunity to fight unfair: the origin story.

Founders love the truth, especially technical founders. They care about precision and detail. They have the sense for integrity that comes from years in the academy. It’s a curse.

In some industries–regulated ones, like medical or supplements–integrity and accuracy is non-optional. But you knew that and have lawyers looking over your shoulder already. For everyone else, it is a curse.

The Harry’s story

Harry’s founder Andy Katz-Mayfield went into a drugstore in 2011 to buy a razor. But the experience was frustrating–locked case, high price, confusing products. His friend Jeff Raider had co-founded Warby Parker, and Andy vented to him. Boom! They started Harry’s.

The Gillette marketing machine has decades of research, hundreds of millions of dollars in marketing, distribution literally everywhere and a global sales team pushing its products. Yet the Harry’s brand and story captured everyone’s attention. In 2019 Shick tried to acquire Harry’s for $1.37 billion (but the FTC blocked it).

From start to FTC-cares-enough-to-sue in under a decade. Not bad for a founding story. It’s a boring listen on How I Built This–no near death experiences, no founder drama. But worth a listen, because boring in this case is The Way.

The opposite of Harry’s

A founder I recently chatted with started his pitch with a great origin story. He was filling out a medical form, and didn’t know what something meant. He had to leave the form and go research it, but it turned out to be an ambiguous acronym. ChatGPT gave him a completely wrong definition, he filled out the form wrong, and it took months to set the situation straight–which in this case meant it took a long time to get medical insurance to reimburse him.

Pretty great story. Real pain, he understood it, and as an AI-inclined founder was well positioned to do something about it. I leaned forward in my chair, expecting some kind of embeddable chat for medical forms AI app.

Instead, he pitched a product for influencers and online course teachers to host websites and have AI interact with their followers. WTF?? Where did that come from?

Do I need to explain what’s wrong here, or do you get it already? He didn’t. I had to explain it. I felt bad because I didn’t think I was being nice. And I kind of wasn’t–that voice in the back of my head was screaming “This guy is wasting your time.” But not telling him would have been far worse.

This kind of founding story isn’t like the Harry’s one. You need a Harry’s style founding story. If you don’t have one, make one up. Seriously. People do it all the time.

The (fake) Houzz founding story

I met a founder of Houzz, the very popular app that dominates home decoration and remodeling media. There were two founders, both developers but one who leaned into business and one who was an iOS developer. The founder I met was the iOS developer. He built the iPad app and a lot of the early stack.

“But wait!” you say. “Houzz was founded by a husband and wife team! It even says so right on their website!” It’s an amazing story! Alon was a developer at eBay and Adi was his wife who was trying to remodel their Palo Alto home! Because it was so hard to find good help they built Houzz to help ideate and find great contractors. Who wouldn’t feel badly for that cute Palo Alto couple? They were on BBC, CNBC, and lots of other places.

Apparently everyone empathized. It is a pretty amazing founding story. Houzz is also a great product, one that many of you have probably used. It has raised over $600 million and was last valued at over $4bn.

But the founding story is total bullshit. Maybe there was a Palo Alto home. I’m sure Alon was a developer at eBay. But Adi didn’t join at the beginning–she joined much later. The guy who was there at the beginning was the developer I met. He built the iPad app that the company was built on top of, long before Adi joined. At least, that’s my understanding of it all.

Does it matter?

No.

No it does not. There are no rules. The truth is what you say loud enough and often enough until it is the only thing people remember.

Don’t feel bad for my developer friend. He owns a lot of shares. He’ll be fine. Better, actually, having been cut out of the founding story than if he had been left in. What would that have been, anyway? “Two nerds figured out a great market opportunity in a rich suburb and exploited it with their software coding skills.” Can’t see the BBC–let alone House Beautiful–picking that one up.

Would Houzz have been as successful if it talked about being two developers with a damned good idea? Maybe, but probably not. The PR story sure sounds better when it’s a husband and wife romcom. And it made PR a major driver of growth.

Don’t be crippled by the academy

The truth may well be that the idea to start Houzz came from 2 developers in East Palo Alto filling out a medical form (in a fixer-upper house?) and needing to look things up–but getting it wrong. The world will never know.

What matters is the story and the impact it has (or can have) on your business. There is no bad consequence for telling a great founding story that might not be entirely accurate.

The consequences for telling a lousy founding story are severe.

Do it right.

Welcome to the latest S2S post. Three important updates.

1. We’re cleaning up our email list. Click a link (any link) so we don’t accidentally unsubscribe you! Seriously.

2. We’re considering a paid tier and/or sponsorships. Would you pay? Would you sponsor? Just reply…

3. Planning an event in SoMa for SF Tech Week–would love to see you there. Want to speak? Sponsor? Host? Just reply…

Peter Walker of Carta released his latest analysis last week. It should be required reading for all founders. Check it out. I’ll focus on M&A, GTM and hiring in this post but there’s a ton more in there worth thinking about, like equity grants, time between rounds, and more.

Carta’s data on M&A

Let’s start with good news. M&A is on the rise in 2025. This matters because M&A is correlated across companies, and that impacts outcomes. If doing M&A at a board level is viewed as not smart, most potential acquirers will not pursue M&A. That means that if you do manage to get an offer, you might only get one. And it will be harder to get anyone to make an offer.

When M&A is viewed as a good idea, it can be contagious. Board members across companies will bring that idea with them, and be more likely to approve an inorganic growth strategy. More companies will be acquisitive and not only is a founder more likely to get an offer, but also they are more likely to get multiple offers–which is how you get the best outcome.

M&A by quarter, from Carta and Peter Walker

If M&A is on the rise, it’s cyclical and likely to pick up even more. You can see the chart in 2021 and 2023-2024–deals beget more deals, until the music stops. If you are considering an exit, now might look like a good time.

Three additional thoughts. First, that doesn’t mean outcomes will be great. You can’t count on taking your $250k ARR startup to market and sell it for 50x revenue. Or even 10x. It just means that board members won’t vomit on the idea of M&A the second an exec brings it up. Which, at times, they do.

Second, there are more startups in general–especially including seed stage, unfunded startups–than there were even 5 years ago. So this shouldn’t be taken as a probability of getting acquired. If anything that likelihood went down, even with more deals being done.

Finally, feels good to know I led ~2.5% of the deals that closed in Q4 2021. Just saying.

GTM is (still) the sine qua non of startup success

Sure, you need a good product and yes, using AI will help you raise and at a higher valuation. But the ARR goalpost has moved–from $1 mm in ARR at Series A to an average of $2.9mm. Now $1 mm puts you in the bottom quartile–a place nobody wants to be. You’re not raising from top tier VCs with a bottom-quartile key metric, unless you (a) are an exited founder with a proven record or (b) you achieved that extremely quickly.

The Series A goal post moved (again)

This goal post has always been moving. In the 20-teens it moved on me mid-raise, or at least it felt that way. I don’t think it’s just inflation. I think there are two drivers. First, people are really dialing in their GTM. They get it, and get it early, and just do everything “right”. If you were an investor would you want to invest in someone who figured out how to win the game, or someone who might figure that out–but hasn’t yet?

Second, there are a ton more startups. Y Combinator is doing 4 batches a year with ~400 startups in each batch. There are something like 5 companies every batch doing what skyp.ai is doing, for one example. Investors pick the one of those five who has figured out GTM. That company will then have a war chest to extend its lead.

The chart does not include gross margins, LTV, CAC payback, or other key business metrics. So while the goal post moved, realize that raising to hit these metrics might be a losing game for you as a founder. Sell 20% of your company for $10 million so that yes, you can hit $6mm in ARR but at the cost of burning $12mm / year and needing to raise a Series B quickly. You may find yourself running to stand still at any moment.

Know the game you are playing. If you are raising venture, you have to play the game on the field. That means if you’re not at at least $3mm in ARR in 2025 raising a Series A will be hard. At $1mm ARR it might be impossible. Lower multiples, worse investors, longer time to get a round done. Plan accordingly! Invest in GTM early–maybe even before your product is done so that when your product is fully built you’ve got a waitlist, or customers, or some running start. There is no GTM channel that takes less than a quarter to ramp up. Most take several.

Later stage ARR

It’s not in here but worth mentioning. If you’ve been at it for 5+ years, raised your Series A back when $1mm was the requirement, and are now sitting at $3ish–you’ve got a problem. You can’t raise another A, for obvious reasons. But you have Series A metrics.

Your best shot in that scenario is to either bet the farm on growth (not necessarily a recommendation–it depends) or to start thinking about M&A. Trim down costs, “run it like a business”, maybe raise prices, and see if you can get an outcome while the M&A window is open. It may not be the outcome you want–but you will like the terms on that not-quite-Series-B term sheet even less.

If you have supportive investors and they want to bridge you or continue to fund you, that is awesome. But really do the math on the dilution. Because…

Dilution is (still) a bitch

While we’ve always known that dilution sucks for founders, this deck really puts it in black and white. Investors own over half the company at the Series B–but founders own less than half after the Series A.

Dilution is a bitch for founders

I’ve worked with founders–YC founders in particular–who raised $3-10 million on post-money SAFEs, over a few raises, without ever doing the math on dilution. They come into an M&A or Series A raise to discover they actually own <25% of their company. Be careful. And note that chart is for all founders–not just 1 founder. So if you have a founding team, each might own a ex-co-founder you fired might still have a big chunk of that orange slice in the chart.

These numbers are medians but are totally in line with what is common. Unless your company is on fire, when more capital drives more growth, raising more equity tends to benefit investors (who can invest more and keep their ownership percentage stable). I mean this in terms of outcome–what you make from the sale of your company.

Take a simple example using the above numbers. You raise a Priced Seed. You then get an offer for $20 million to acquire your startup. This is only ~5mm above the valuation at the last round (of $15mm). But with the ownership of 56%, founders take home $11.2 million. Not bad. Perhaps they turn the $20mm offer into a competitive process and get an exceptional outcome: $50mm. In that case, founders would get $28mm. Either is a great outcome for a founder. Obviously the $50mm one is better.

Let’s say that they turn down the $50mm offer and press on. This happens all the time – Maxiumus Greenwald recently posted that Warmly turned down an offer to press on with growing a big business. There are stories aplenty of venture investors refusing to sell. Digg, perhaps, the most famous recent one–where Newscorp offered $60 million plus a $20 million earn out, but the board refused. Things went south from there.

Back to our example. Let’s say the founders press on, and get through to Series D. I’ll try to avoid complexity, but the average Series D was ~$65mm in Q4, 2024 with a premoney of ~$600mm and for this example, let’s assume Series D investors get participating preferred rights.

A few years later, the company gets an acquisition offer for $665mm, the post-money valuation of the last round. At this point the founders are tired, ready to move on, and the market has changed. There’s no obvious path to a better outcome, so founders and the board agree to take the offer.

Because it is participating preferred, the D investors get their $65mm back off the top, so $600mm is available for distribution (and the Series D investors participate in that, too). The founders get 11.4% of $600mm, or $68.4 million. Yes, that’s better than $28mm. But how much time has gone by?

It was 7 years between that priced seed and the Series D. Nobody sells for exactly what they raised their last round at immediately–they struggle longer. Years longer. This offer was not the ideal scenario, it was another 3-5 years of struggle later so roughly a decade since that first offer.

If the founders had just taken their $28mm a decade earlier and invested at 7%, they’d have $51 mm. Not to mention any retention equity or juicier big company salary or benefits. And a better golf handicap.

This example is flawed because it focuses on a pretty good scenario. Plenty of Series D companies don’t make it–they get sold for $200mm. Or $0 million. At that point, the founders get less cash 9 years later than they would have if they sold early. But founders are an optimistic bunch, myself included. You don’t plan this out thinking “mid- to worst-case”. Maybe you should? The mid outcome early on ($20mm) sure beats most mid to worst case outcomes later.

Other thoughts for later posts

The data on hiring and employee grants is interesting. Basically–employees don’t get much equity anymore, after the first one. Hiring is also very slow. These may be related. Perhaps employees are joining when there is far less risk and the business is further along. Why give someone 5% if the business is already established?

There’s something else in that data. Nobody gets to $2.9mm ARR with zero employees. Well, maybe a few do, but it’s rare. There are either a lot of founders (Anthropic had 7) or contractors working without equity. I’ve noticed a lot of very successful “3 person companies” actually rely heavily on expensive and well staffed outside agencies.

Peter also talks aobut the AI trend towards fewer employees. The AI trend feels real. Yes, hiring fell because funding fell off in 2024. But now it feels very much like more is getting done with AI. That doesn’t mean zero hires–it means more is expected of each one. Why hire a marketer and a copywriter when you can just hire a marketer? Why hire an engineering manager and two engineers when you can just hire an engineering manager and zero engineers? Anyone whose job was to write data queries should feel very nervous: it’s far easier to ask ChatGPT to write a BigQuery query and iterate on it than it is to ask a human, who is possibly in a different time zone, off at lunch, or working on someone else’s query request.

I think we’re shifting to hiring people who know how to play the game, and then using AI to actually play the game. Versus needing both people in leadership positions who know the game and people in player positions to play the game. The implications go far beyond startupland.

Some conclusions

Figuring out GTM is critical early–and at all stages. You have more time between rounds to get it right, but the bar is much higher. There is more noise and more competition.

Dilution comes at a greater cost than ever. If you have a favorable offer early–when you own half or almost half–take it seriously. It’s exciting to read about so-and-so’s $500mm exit in TechCrunch, or picture yourself bragging to your fellow YC classmates (yes, it is bragging) but a $50mm exit now might actually be a better outcome for you even on cash terms, let alone the time value of money. And of your life as a founder–resting and vesting is easier than the grind.

We did not need Carta data to know this. This is not news. Mark Suster talked about it in 2011. But the Carta data makes it clearer than ever.

One thing I haven’t touched on much is alternative financing. There are new funding methods for later stage companies that can massively lower the cost of capital while increasing growth rate. Perhaps I’ll get to those in a future post… Any interest? Drop me an email.

One reader and good friend pointed out that after writing on how to sell your startup, the logical next post should be on how to optimize your exit. That means how to pay less (or no) taxes on your sale.

There’s something about being a second-time founder that makes you think about optimizations earlier. Maybe it’s all the taxes you paid on the first one. Maybe it’s just experience.

QSBS might be the most important one. If you’re a repeat founder, the only part that matters is what’s changed under Trump. If this is new to you, maybe gift me some stock later as a “thank you” or “attaboy bonus.”

How to sell your startup and not pay taxes

Here how you sell your startup and not pay taxes, the way Steve Martin would present them:

• First, build a business that is valuable (growing fast and, even better, profitable).

• Second, sell it.

• Then, don’t pay taxes.

• When your accountant says, “Didn’t you sell your company for a pile of money?” You say one simple acronym: “QSBS”. And that’s it!

Ok, that’s not all of it. Let’s dive in. Here’s a short summary of QSBS and two other tax mitigation strategies.

What is QSBS?

QSBS stands for Qualified Small Business Stock, and is a way the tax code incentivizes investment in small business. It applies to founders and early investors. If your investment (as an actual investment, founder shares, or conversion from an LLC) meets certain criteria, you do not pay federal taxes. You may still have some federal tax liability and you may have some state tax liability (California, especially–but that’s out of scope). But none from the federal government.

This is about taxes so it’s necessarily a little wonky, but I’ll try to keep it straightforward. The two things to take away are:

1. Pay attention to QSBS qualifications and make sure your ownership of your startup, and (if possible) your investors’ ownership qualifies for QSBS.

2. Consider taking active steps early to maximize your QSBS impact.

The “Big Beautiful Bill” signed on July 4 increased a lot of QSBS limits and made more sales qualify for QSBS treatment. Why this giant handout to founders made it into the bill is a phenomenal question, but out of scope for this post (or newsletter). If you know, I’d love to hear–but for now, let’s stick to what you can do to take advantage.

Disclaimer

This is not tax advice. It is 100% worth talking to a professional about this if you are a founder about your specific situation. I’ll give you the basics–and even the advanced tactics. But this is just for your entertainment. Read on, young founder, and profit.

The basics–qualifying for QSBS

To qualify for QSBS, the following requirements must be met:

This might seem complicated, but for an early stage tech startup founder or angel investor, you probably meet all of these QSBS requirements without trying.

If you don’t do any optimization, but do qualify, each founder has a $15 million tax exemption if they sell after 5 years, and part of that starts after just 3 years. Who wants an I-banking bonus or truckloads of Google RSUs taxed at 37% when you could pay… drumroll…

0%

Don’t mess it up

The exclusions are important. Secondary sales are excluded, but not gifts. How you hold the shares matters. Trusts are fine. Many other entities are not, so talk to an advisor.

Some excluded business types are tech-adjacent, but you might be able structure two separate entities to get around this so the tech lives in a qualified entity while the unqualified aspect of the business lives in a separate entity. Again, get pro advice.

Advanced strategies for QSBS

There are two popular advanced QSBS strategies that are proven. They are called stacking and conversion.

Stacking is more straightforward. The $15 million cap applies to each individual shareholder. So if you have 1 founder, that’s $15 million. If you have 2 founders, that’s $30 million. If each founder gave shares to 1 child each, that’s 4 shareholders or $60 million.

To stack, early on–before the company is worth more than the $75 million gross assets limit–gift stock to other qualified people or entities. (Yes, this went up from $50 million with Trump’s BBB). For example, you could gift it to your child’s trust. Or a relative (probably not–but maybe, your spouse). Again, talk to your tax professional about your specific situation as there is some uncertainty in IRS policy.

Conversion is more complicated, but in some ways cleaner. The (new) $15 million limit is great if you sell your company for, say, $15 million. Or if you own 15% when you sell it for $100 million. Either way, you don’t pay federal tax.

But what if you sell it for $600 million?

If you are using the $15 million limit, you would owe tax on $585 million in gains. Even if you handed out shares to family members like Halloween candy, do you have enough children or family members to shield the remainder? Doubtful. Also, obviously, giving it away means you (might) get 0 benefit rather than whatever is left over after the taxman takes his bite.

Hence, conversion.

Instead of incorporating as a C-corporation and qualifying immediately, let’s say you start as an LLC. You build a nice business. It’s starting to take off, but before it really gets going you convert it into a C-corporation. You read this article, and have good tax advisors, so you do that when it’s worth $60 million–comfortably below the $75 million cap.

Your QSBS exemption applies to 10x your investment, which in this case is the $60 million. When you sell for $600 million, you would be liable for taxes on the $60 million you converted–but exempt on the remaining $540.

That’s right: you’d pay federal tax on ~10% of the sale. And pocket half a B tax-free. Well, at least federally tax-free.

Would you leave California and not come back for $77 million? (A lot of people answer “yes” to versions of this question)

There are only three downsides to this. First, the clock resets on the conversion. So if you ran your business as an LLC for 3 years, and then converted, you’d need to wait another 3-5 years to benefit from QSBS. Second, if your company takes off and is worth over $75 million, you cannot get any QSBS benefit (because the conversion would not qualify). Seems like decent risk-reward to me, but with the increase to $15 million it may not be worth it. Third, if you are reading this post after you’ve already started a C-Corp then that ship has sailed and you can’t use this strategy. There’s always next time.

The biggest way to mess up QSBS

The best way to mess up QSBS for everyone is to not build a valuable business, or to not get much for it if you do. You can optimize the shit out of your tax situation, and then not get a deal done – and get $0. Focus on customers. Build a great business. Then run a good process. Maybe it’s suboptimal, but paying taxes on a life-changing windfall is better than paying no tax on a total loss.

A great way to mess up QSBS for your early employees is to not let them early exercise. Yes, I’m still bitter about this. You know who you are. To benefit, the holder must hold the underlying shares (not an option) so let your employees early exercise and do their Schedule 83b election. If they leave, you can always buy the shares back at the strike price. I have a post in the hopper about reducing your admin work, but this admin work is 100% worth it.

Another great way to blow it is to take your eye off the ball during the actual M&A process. Many of these optimizations will be difficult to execute concurrently with a transaction, and could delay a transaction. Time kills deals. If you sort tax optimizations out early, you will have one less distraction. And some strategies require being set up early.

Other tax strategies

There are other, more aggressive tax strategies for US founders which may be worth talking about with your advisor. Both require some advanced planning and should not be done concurrently with or even in close proximity to a sale. One may be quite risky.

The first, risky strategy involves putting your ownership in an offshore trust for a non-US beneficiary. This is known as an offshore trust or foreign grantor trust (FGT). The goal is to move your assets out of US tax jurisdiction. Because the US taxes its citizens and permanent residents on a global basis, the only way to do that is to not own the assets.

This seems extremely risky to me on two basic levels. First, you are giving someone else your stock. That might be a close relative in a foreign country, but fundamentally has the potential to go badly. I’m sure lawyers have thought this through, but still. Second, the IRS is hip to this strategy and frowns upon it. If it determines that the beneficiary is a “straw man”, it can pierce the veil of the trust and tax you directly. I am not a tax lawyer or CPA, so if you are curious talk to someone who knows about this if you are considering it.

The second is a broader estate tax planning strategy, in which you set up a life insurance trust with grantor trust features, and do a lot of complex things beyond the scope of a simple newsletter. Again this is the purview of very advanced tax and estate planning, not for your everyday founder, but if you may receive a huge windfall–and if QSBS isn’t going to help you much–it’s worth talking to your advisors.

Both strategies incur significant up front cost, so if your company ends up being worthless and you’ve spent $20k+ to tax-optimize, well, that sucks. Which is perhaps why its mostly second-time founders who do this sort of thing–they have the money to blow. And are probably a little bitter about the tax bill the first time around.

QSBS has the advantage of requiring little to no extra work, time, or expense, while providing significant tax optimization–at no risk. The IRS guidelines are quite clear, and uncontroversial. After all, the supposed purpose is to encourage small business investment. Consider other options if you must, but at the very least–ensure you’ve got QSBS set up properly.

Good advice and the best advice

I talked to an entrepreneur this week who had never heard QSBS even though she was almost done with a fundraise. Shame on her lawyers and accountants. Make sure your advisers know their stuff and are, as Edith Harbaugh (LaunchDarkly) says–”stage appropriate”.

Take action early–not only to be able to focus on closing your transaction when the time comes, but also because some of the strategies mentioned need to be done early. But do so with experienced, professional advice! Too many zeros involved to DIY or trust ChatGPT with.

Perhaps the best advice? As Steve Martin said, explaining how you can be a millionaire and not pay taxes: “First, get a million dollars.”

Don’t forget–that part comes first.

Welcome to Seed to Sequoia! Sending late today unapologetically. I realized in writing the Windsurf post that I never wrote a “typical” how to sell your company post. Here it is–hope it’s useful.

If it is useful, please tell me. And if there are topics you want to hear about, let me know! There’s so much to cover in startup strategy, GTM, and M&A it’s hard to know what to cover. Just reply. I read and reply to every one.

Nobody tells the truth about a acquisitions in public, so it’s hardly a surprise that many founders make M&A decisions based on partial or simply wrong information. Much of the advice out there is bad. Or worse, it’s based on TechCrunch articles.

This is both the best of times and the worst of times to be a founder. Many companies are in survival mode. Not enough growth to raise the next round. Burned all the bridge rounds they could get.

Some founders will want to persevere. But many founders will want to move on with their lives and do something else. They might hire a CEO, but at a sub-scale business, that may not be an option. So they choose to sell.

M&A advice

If you’re considering M&A, sure, read this article. Then get advice from a pro. Could be a board member, could be an investment banker or other advisor (like, possibly, me). No founder wants to go to their board and get basic advice on M&A. You want to go to your board with confidence, options and plan in hand. So here are the basics. It’s worth knowing this stuff even if you don’t plan to use it for years. Then get advice specific to you. Can’t say that enough.

Timeline: how long does it take to sell a company?

Most processes take at least six months–but some take years. This is because companies are bought, not sold. You run a process to find buyers, but just because you think a company should buy your startup doesn’t mean they will. The humans at that company have to decide to buy it, for whatever their own reasons are.

As a result, the timing is unpredictable. The minimum is three months–it simply takes that long to have conversations with possible buyers, negotiate terms and do diligence. But this is accelerated and rare. Plan on six months.

In a worst-case scenario, the process can take 18-24 months. You might need to clean up accounting, IP, or financial structure. Or a transaction might require regulatory approval. Know your specific situation–your lawyers can probably give you some guidance, or an experienced board member or–I’ll say it again–talk to an advisor.

Never expect to sell quickly, even if suitors have inquired. Asking if you’re interested in selling is free and takes no time at all. Actually getting a deal done is something else entirely. Lots of founders get excited when a corp dev person asks about M&A only to discover later that they were just making conversation, and had no intent or mandate.

What if I’m almost out of cash and don’t have 6 months?

You can still snatch victory from the jaws of defeat, but you have to act quickly and decisively. I’ve seen two founders get great personal exits (both to eBay) after basically crashing the plane. They walked away like old-school James Bond, without even a scratch – and a few million in the bank. Which, by the way, I’m sure James Bond has squirreled away in a Swiss account somewhere.

If this is your situation, I wrote a post for you: selling when you’re out of runway. If you’re considering shutting down, there’s also a post about that. Highly recommend knowing all of your options and really digging deep to figure out what you prefer, rather than what you think you have to do.

Preparing for an M&A process

If you’re considering an M&A process, and have time, you can do a lot to maximize your outcome. You want at least three well capitalized bidders. Like a venture fundraise, you want to receive those bids at the same time (hence the word “process”).

If you have enough revenue (or profit!) you can retain an advisor or investment bank. Typically banks and advisors will only work with companies they believe they can sell, and part of the value bankers bring is warm relationships with buyers. Strategics and financial acquirers tell them what they are looking to buy, so if your company fits with something people are looking for–they will make quick money working with you, and will be enthusiastic.

Most of the banking and advising business is based on success fees. There is little incentive for an engagement where success is unlikely. It’d be like hiring a real estate agent to sell an unsellable house: they’d simply decline to represent you, because it will never pay.

If you have never sold a company before, know that there are fewer higher stakes skills you could try to learn on the job. Would you rewire the electrical systems in your house or hire an electrician to save $10k?

If you don’t have a board member or other heavily invested person with expertise, it’s worth getting professional advice. You might begrudge an advisor’s 3-5% fee, but if you get $0 because a rookie mistake by you or your team screwed up the deal, 5% starts to feel pretty affordable. The warm relationships with buyers are also valuable, and can help get deals closed.

If you engage a banker or advisor they will give you specific guidance. You can probably stop reading now, and just do what they say. While bankers are great and highly effective, always remember that they are your agents–not principals like you. They get paid on a sale and (usually) only on a sale, so like any agent, their motivations are different than yours. They might do 8 deals a year–you’ve only got one, if you’re lucky.

Running the process yourself

If you cannot engage a banker, or simply choose not to, you will need to run the process yourself. This is similar to running your first institutional fundraise, but differs in a few key respects. The people are professional buyers-of-companies, not VCs. Their expectations are different in every way. There will not be another round, so buyers will be even more motivated by FOMO. Diligence will be more thorough because the acquisition will go on their books, liabilities and all. And the offers can be wildly different–which perhaps isn’t that different from a venture round, but means that one company might offer equity, another cash, a third a mix, all at very different valuations and with different non-financial terms that will affect your day-to-day. Whereas VCs all offer cold, hard cash.

Running a process is a major distraction to running your business. If you’re late stage enough to have a CFO, or someone with the skill to run the day-to-day on your team, this may save your business in the likely event the deal falls through.

The process usually conforms to the same linear steps. You want it to conform, or you’ll lose control. These steps are:

1. Get a board mandate

   If you have a board, you start by getting their mandate to run a process. Understand what their expectations are, and help guide them to realistic ones.

2. Prepare the materials
   Build a tight 1-pager and presentation. You should have a story, but maybe exaggerate less than you did for your Series A. Clean up financials, IP assignments, contracts, and cap table. Make sure you will hit your forecast. Nothing kills a deal like missing targets during diligence.

3. Build your buyer list
   Identify 10–30 targets: strategic acquirers (competitors, partners, platforms), PE rollups, or growth-stage companies with product gaps. Ideally you already have relationships with 5-10 of their CEOs.

4. Run a quiet outreach
   Use trusted intros (board, angels, advisors). Maintain control—start with “exploratory conversations,” not formal process language. Ideally you started this journey with an offer in hand, but exploring options makes it easier to save face if it goes nowhere.

5. Gauge interest & shape deal contours
   See who bites. Early feedback helps anchor valuation, deal structure (stock vs. cash), and whether it’s an asset sale or equity deal. If nobody bites, stop. You will not convince them however much you may try.

6. Create competitive tension
   If multiple buyers engage, set a soft deadline for indications of interest. Even informal competition increases leverage. A tight timeline prevents the process from dragging and distracting you from running the business.

7. Negotiate LOI
   You might receive multiple LOIs. You can negotiate all of them, simultaneously, until you sign one with exclusivity (which means you have to stop talking to other suitors). Focus on valuation, structure, retention, earnouts, and non-competes. Don’t skip legal review—LOIs often include binding terms (exclusivity, confidentiality). Exclusivity has been used to kill companies–be careful. But it is standard.

8. Diligence
   Be ready for 2–6 weeks of deep scrutiny: financials, IP, legal, team. A clean data room speeds this up. During this process you do not want to (a) unearth surprise skeletons or (b) miss your sales or other metrics. Both can lead to deals falling through, or the key terms being renegotiated.

9. Negotiate the definitive agreement
   Watch for traps: reps & warranties, escrow, clawbacks. Retain experienced deal counsel. Expect back-and-forth before signing. This can get very complex–for example, if it’s a windfall you might set up trusts for your children, trusts to maximize QSBS, retention for employees, vesting, etc. Deal fatigue is real–don’t succumb. Stay focused. Your buyer does this 5-10x/year. You do it 1-5x/lifetime.

10. Close and communicate
    Manage internal morale and external narrative. Announce only after wires hit (or when legally required). Lots of deals fall through at the last moment (I’m thinking of 3 right now… one company survived, the other two did not).

11. Integrate
    Merge the companies (or products or assets). The first six months are crucial, and most of the time management will be required to stay at least that long. Depending on your stage, and the deal structure, you might be incentivized to stay for longer–two years is typical, four is not uncommon but is on the high end.

It’s intense–for months. Do not expect to get much other work done, especially when you get into diligence. And many deals fall through, so get comfortable with doing a lot of work and having nothing to show for it. This is why secrecy is usually the best policy, even if up to that point you’ve run a transparent company.

The flow of money

In most deals you can expect to see payments at 3 times:

1. At the time of closing, wires are sent. If it is an equity purchase, all debt holders and shareholders would be paid as would any advisors (investment bankers, accountants, lawyers). 0-15% of the deal value is usually also put in escrow at this time, to cover liabilities missed in diligence or that arise later.

2. 12-18 months later, money in escrow is paid out. Note the amount escrowed is negotiated.

3. Earn outs or other incentives are paid over time, depending on the deal terms. Sometimes these are worth much more than the initial payment, such as in an aquihire.

An asset purchase or other structure might delay payments, because the initial wire goes to the company and then the company (typically) winds down and wires the remainder to shareholders and whomever else it owes. The wind down may coincide with the closing or it may be months (or a year) later, to protect the acquirer from lawsuits.

Avoid loans from the acquirer

I have heard VCs suggest this as an option to cash-strapped founders. It is an option but it is a terrible option. Any acquirer who gives you a loan will put a lot of protections on that loan–to the point where, if the acquisition falls through, they may end up owning your assets and company anyway. This sort of thing slows and complicates deals, and can be avoided simply by managing cash carefully and starting the process early enough.

A better option, if you have it, is to tap an existing line of credit from SVB or a similar institution. This can also be outrageously expensive (one company I acquired spent $50k to borrow $100k for 1 month) but it’s better than the alternatives.

Advice and how to play it

Every process is a little different, and there’s skill in running an effective one. If you’re considering an acquisition, get advice from someone who is experienced. I’m one such person (just hit “reply” if you want to talk about this). There are many others out there–some who may have specific insights into your industry.

Investment bankers are a great resource. There is no harm in building relationships with prominent niche bankers in your industry. This will put you on the radar of possible acquirers earlier, and also give you insights into roadmap if you want to steer towards a gap where acquisition is more likely. Don’t, however, lose sight of customers or their needs.

That is the main risk to M&A: it can be a distraction. If you succeed it might be life-changing, but success is not guaranteed. Having an advisor or a CFO who can manage the process and allow the rest of the team to continue to build and maintain the business is crucial. While I personally think being transparent is the best way to run a company, secrecy around M&A is required because employees (and even investors) are ill equipped to handle the anxiety that it can create.

If you can’t execute on M&A, you can always shut down. There’s nothing wrong with that choice. This is a hits business. If it’s not a hit, move on. That’s what VCs do and founders should do it too.

Hopefully this helps guide your expectations of an exit. If you have specific questions or experiences to share–please reply. I read and respond to every email.

Enjoy this newsletter? Please forward it or tell your friends! I’m always looking to connect with more founders and help them on their journey. If you or a friend needs M&A help, happy to chat. Find a time here or just reply.

Today’s post is the second in a row on current events–because I think this one really matters strategically to founders, especially early stage in AI. Unless crazy stuff happens, expect to go back to GTM-focused posts next week.

Speaking of GTM, if you’re curious what’s going on over at skyp.ai check out the site–added 10x more content (thanks to AI) including a vibe-coded video series player that took about ~15 min (but I still can’t stop the auto-play). More next week. For now, Google-Windsurf!

The Windsurf deal may seem like just another AI headline with 9 zeros: Google acquires [ AI company ] for [ x ] billion dollars. But to those of us buy and sell startups, and advise on acquisitions, it is a massive change. Here’s what happened, and why it may matter to you.

What happened with Windsurf

Windsurf was in talks with OpenAI to be acquired for $3 billion. The deal fell through. Or was never real. Who knows? Google stepped in, and in a very unusual deal, “acquihired” key team members and a non-exclusive IP license for $2.4 billion, leaving the entity, many employees, and actual IP behind. Days later Cognition AI, the company behind Devin, the $500/mo AI-developer-agent, bought the entity and (presumably) took on its left behind employees.

I don’t have insider info on this (if I did, I couldn’t talk about it) and there are many reasons why nobody will tell a real acquisition story in public. But the broad strokes of the deal are public–and the implications are profound.

The M&A basics needed to follow the rest of this post

Before we dive deep into pros and cons let’s provide context. Here are the three main ways to be acquired:

1. Stock purchase. The acquirer buys all of the target’s stock. This is what “acquiring a company” has traditionally meant. QSBS, capital gains and other tax advantages galore. Note that an actual transaction may look much more complex but fundamentally and crucially money goes to the owners of the stock, not the company (matters a lot for tax).

2. Asset purchase. The acquirer buys the assets of the target, but not the stock. It might get other rights. The startup Shift acquired Fair Technologies’ dealer marketplace assets (code + customer contracts) in 2022, for example. Limits liability but generally accomplishes much the same thing as the stock purchase. Usually not tax advantaged.

3. Waiver and Release. The acquirer buys and owns nothing. In exchange for a payment to the company, it receives a non-exclusive, perpetual and irrevocable license to the target’s IP and the target waives any right to sue it for hiring away its team (or, usually, the right to sue about anything else). Many acquihires are done this way because it’s faster, cheaper, and safer for reasons we’ll discuss.

Google seems to have used the waiver and release structure to acquire what it wanted from Windsurf. 🤔 

The worst way to get acquired

Founders hate the waiver and release structure. It is not, in fact, being acquired at all. Can you post to LinkedIn “My company was acquired?” when it in fact was not? People do it all the time. But then… therapy? Guilt?

The waiver and release structure is to M&A what the reverse merger is to going public. That is: it’s for losers. And yet, the SPAC craze turned the reverse merger into the hottest financing vehicle. Because SPACs have advantages. The waiver and release has advantages too.

Before Windsurf, most companies contemplating this structure did so because they had no better options. But under the right circumstances, it’s a great structure for founders. Like where the value of the knowhow vastly outstrips the value of the actual IP or customer contracts.

The only negative that really matters is that it is tax disadvantaged. Typically the acquirer will pay the company directly to cover the costs of shutting down the remaining entity, and in exchange for the rights in the waiver and release agreement (consideration is required for it to be enforcable). Whatever is left after the company’s liabilities are paid off and is distributed back to shareholders in the shut down process. Usually this is zero, or less than what they invested. Sometimes they might get all of their money back, but that’s rare.

Because it’s usually a fire sale, money usually goes to investors high in the preference stack. There is rarely anything left over for founders or employees. Those people might get a carveout, signing bonus, or equity in the acquirer as part of an employment package but–from a tax standpoint–it’s all bad news. No QSBS, no capital gains. Just ordinary income. Because anything founders or team do make usually comes in the form of future compensation (equity in the acquiring entity, earn outs or targets, etc.) that’s one more reason why this is considered an undesirable structure. But there are reasons that can be a positive–and Windsurf highlights them beautifully.

The silver lining

There are huge advantages of the waiver and release structure–and those advantages center on speed and flexibility. Which sounds like what happened at Windsurf.

Ludicrous speed

The acquirer is just hiring the team it wants, and getting a license to the IP they might or might not use (more as a protection against future lawsuits than because they’ll use it). This means the acquirer doesn’t even really need to look at the books. Or comb through HR matters. Or scour agreements for unlimited liabilities. Or diligence the users. Remember the HBO Silicon Valley COPPA scene? Hooli realizes its mistake far, far too late. Waiver and release avoids that. None of that transfers, and the risk of a lawsuit forcing it to transfer are very low. Super fast.

Waiver and release wins for speed in integration, too. In other structures, there’s typically an integration period of 6 to 24 months where you integrate the systems. Instead, this looks more like a new-hire onboarding timeline. No CRM integrations, no right-sizing SaaS tools, no migrating from Azure or AWS to Google Cloud. Just here’s your new laptop and email login; let’s go!

The speed benefit applies to regulation–which, maybe, deserves its own headline. Because Google wasn’t acquiring any IP or technology exclusively, they avoided any regulatory review. Instead they dodged what could have been a year-long regulatory process. I haven’t heard of this structure being applied for this reason before, but I’m sure it will be again.

Google obviously valued speed. The AI world is evolving fast. Companies like Lovable went from $0 to $80mm in ARR in months. Don’t think growth is slowing, either. So waiting many months for integrations, or regulatory approvals was unappealing to deal-breaking. Instead, Google got a world class team on Gemini relatively instantly.

Flexibility and great power

The other aspect is flexibility. When you acquire a company via a stock purchase, you (generally) pay everyone according to their rights in the cap table. This can pay people who you don’t want to pay or who add no ongoing value to the business.

Here’s a simplified example. You have $100 million to buy a company, and you really only care about 10 key founders and employees. VCs own 80% and a founder they fired 2 years ago owns another 5%. Out of your $100 million, that leaves $15 million to split between the 10 you care about. Worse, they get it based on their stock–so that recent, 100x hire who joined later might own 1/10th of what someone who joined earlier but isn’t as good owns. Some carveouts and exceptions are possible, but challenging. It’s all suboptimal–from a going forward perspective.

The waiver and release throws out the cap table and lets the acquirer do whatever. In the Windsurf deal, it sounds like Google paid about 50 people–founders, key employees, and VCs–the bulk of the $2.4 billion headline price. Rumor is that the 200-250 employees who were left behind didn’t get anything, or next to nothing. If they hadn’t vested stock, which many if not all had not, they weren’t really owed anything.

This flexibility is great for founders. Founders usually have a big hand in who gets the money, and are incentivized for the future–they want to show that this was a great move by the acquirer (so they can brag on LinkedIn?) and also earn their earn out or incentives. So would they give that co-founder they fired $5 million? No. They’d give it to that 100x employee who’s key to the next chapter.

What’s changed

Waiver and release was the reverse merger of M&A. Undesirable. Embarrassing, even. But now Google-Windsurf uses this high discretion, founder-friendly structure in a 9-figure deal. Google was able to achieve its goals of retaining the key team members to work on Gemini products while closing the deal quickly and avoiding regulatory delays.

This can be hard for traditional M&A people to understand. Typically fiduciary duty matters in M&A. It doesn’t matter if you vibe with one company or another; whether one company uses Google Apps or AWS and the other is still on Eudora and bare metal. Whoever pays the shareholders the most wins. Period.

The idea that the founders (and team) choose where they end up based on vibes, or whatever–anything but who paid the most–was a foreign concept.

Now, it seems, it’s entered the mainstream. I bet we’ll see a lot more of this. It’s probably good for founders. It’s probably not great for investors. The people the most screwed are the workers–the sales, the ops, the HR, the workers who in the age of the Facebook or Google IPOs became well off if not actually rich. Now they might be lucky if they see anything at all. From a $2.4 billion deal. At least they can ask AI to help lawyer up for $20/month?

If you enjoyed this and haven’t subscribed you can subscribe here. Please share with friends and colleagues–we’re actively working to growth the S2S subscriber community so that we can take over the world. Or something.

Base44, a vibecoding tool, sold to Wix for $80mm, allegedly 6 months after starting. Making this story even more amazing, Maor Shlomo was a solo-founder operating it solo for the first ~4 of those months.

I haven’t spoken with him, but I was curious so I dug around. Lenny did a podcast on it which is maybe worth a listen (on 1.3x). Those are the facts that are out in public. I have doubts, or at least questions, but let’s just say that it’s all true and Maor really did bootstrap a company to an $80mm outcome in 6 months pretty much all by himself. What can we learn–and what conclusions would be dangerous and probably wrong?

What we can learn (aka What we already knew that Base44 helps “prove”)

1. One person can build a company to multi-million ARR.

2. Big markets mean there’s lots of room for success. Bolt, Replit, Cursor were all direct competitors. They’ve raised tons of money. But there’s also tons of demand.

3. Knowing your customers is crucial. Maor spent the first weeks at a desk with 3 friends watching them use the app and then fixing stuff.

4. The faster you get bacon (what you want) after you push the button (sign up) the better. Maor talked about removing a “review” feature because even though it led to better code it reduced conversion because it put an extra step in the path of getting the “aha moment” (his words)–or, bacon.

5. The economics of selling without any VCs (or anyone else) in your cap table are amazing.

6. Running a profitable business gives you a lot of options.

7. Just because the “big guys” are running $1mm hackathons (insert any big-budget marketing ploy here) you can compete with a $5k hackathon “for good” (insert any targeted, narrowly focused and well executed marketing ploy).

8. The hardest thing is figuring out the go to market.

9. You have to spend time on go to market, even if you’d rather be coding/shipping features/doing something more fun.

10. Automating your own workflows using vibecoding can dramatically increase efficiency and the output of any one person, in almost any discipline (marketing, ops, engineering, product, etc.).

What would be dangerous to conclude

1. One person building a company is the best strategy.

2. Not raising VC in a big market is the best strategy.

3. You or anyone else can actually achieve this kind of success in 6 months.

4. Being acquired early on a hypergrowth trajectory is a good long-term strategy.

5. Their main growth path, building in public via LinkedIn, will work for you or anyone else. (Maor happened to be selling building so building in public appealed directly to his audience).

The most important conclusion

What is abundantly clear is that Maor found something that a certain set of people wanted, built it (quickly, apparently) and they became massive advocates for it. And, of course, paid for it. So many founders–bootstrapped, well funded, across the spectrum–forget this part. You’ve got to build something people value.

I was reminded that setting yourself apart is critical. Having a different looking website, language, messaging, the rest of it is important. But more important is standing for something different–something customers care about. In Base44’s case that is “batteries included”: backend built already, not with AI but in a deterministic way that works for sure and is well architected (and hosted by them, not third parties). If you’ve ever vibe coded, these are real frustrations and a real problem worth solving. Apparently Maor solved it.

Echoes of a past miracle conversation

Around 2010 I met the founders of Braintree, the Stripe competitor PayPal acquired for $800mm back in 2013. When I asked how they got started, they said they were just a bunch of people (40, to be precise) working on payments tech in an office in Chicago. And it “just took off”.

Bullshit.

Maybe if you were 3 or 4 college roommates, I’d believe you. But 40 people? That kind of accident doesn’t just happen.

I get a whiff of that here. Not the 40 people–just the incredulity. I would love to believe Maor built this all in 6 months, start to finish, working by himself for most of it. But it just sounds like such perfect, amazing, zeitgeist-aligned PR that I have questions. Doubts.

That doesn’t detract from the achievement–I just think it may be a disservice to the rest of us who, what, fire our teams and login to Base44 to build everything themselves–and market it?

Thanks for reading! I’m teaming up with Gregory Kennedy of VibeYourSaaS for a live webinar on 3x’ing your outbound. Register now–and, if live events are something you’d want to see more of, please reply and tell me what would be interesting.

Over the last couple of months building skyp.ai and talking to literally hundreds of founders about their outbound, I feel like I lost the plot. Microcampaigns are where it’s at. Here’s why, and how to run them.

I lost the plot because the current AI world has taken scalability to a whole new level. What used to be only for enormous, well funded companies with dedicated growth teams has become common at any company, especially startups. Whether stringing together automations using vibecoding and OpenAI’s API or leveraging purpose built products like Clay or UnifyGTM, automating GTM strategies has never been easier.

The rise of the “AI SDR” means you can run “plays” all the time, emailing literally everyone all at once. Some of these are effective (most are not) but the feeling that you’ve solved the problem by articulating a strategy, automating the shit out of it, and having it always running in the background is a great feeling. Even if it doesn’t really work. Maybe you’ll learn something.

As a startup, however, these scaled “plays” are not the way. Because they are automated, they take a lot of thought and setup. They require significant budget. They can flood inboxes and calendars with unqualified leads. Usually they don’t work all that well, so you make up for that with volume. In small markets, or for early stage, targeted products, scale isn’t possible.

Microcampaigns are the answer. They solve two major problems for startups. First, they get results–qualified people to talk to and sell. Second, they enable experiments.

What is a microcampaign?

Microcampaigns are targeted campaigns to a very specific set of people. You might reach out to speakers at TechCrunch Disrupt. Or attendees who fit specific criteria. Maybe you target leads who took a meeting last quarter but never bought. You might run a campaign targeting portfolio company founders of a specific VC (which, now that I write it down, sounds like a pretty clever idea that I should try).

While microcampaigns could be executed in any medium–email, sms, paid search, display, outdoor, etc.–usually they start with or at least include email. This is because email at small scale is very fast to get going, if and only if you keep volumes small.

Any specific micro campaign is usually a one-off. You are not always reaching out to founders debuting at Disrupt. It only happens once a year so you do it just the one time. Every year, if it works. But the strategy of running microcampaigns is evergreen. You can (and perhaps should) always be trying new campaigns to see what works.

What makes microcampaigns interesting

Microcampaigns simply work. The nature of the campaign itself makes it effective. A microcampaign speaks specifically to a very narrow audience. You’re not trying to reach every startup founder; you’re talking to ones with, say, a board member from Congruent Ventures. Or a booth at TC Disrupt. The more targeted, the better. This makes everything about the campaign perform better.

The homework necessary to build a microcampaign leads to its success. Think about the work required to find portfolio companies of a particular VC. Even if you outsource the actual research to ChatGPT you can’t avoid learning something, spotting some patterns, and using those insights in the campaign.

Microcampaigns do not have to be email. Delve, a startup doing compliance, sent custom doormats to other recently funded startups. Obviously you’re not going to send 10,000 doormats. But they sent around 100, and got a lot of meetings.

That said, email based microcampaigns outperform. We see microcampaigns get 70-80% open rates and 10-20% click rates. Many event-based microcampaigns can see a 10% or higher meeting rate, with interesting and qualified leads. Compared to industry leading AI outbound campaigns in the 0.5-1.0% meeting rate, it’s worth at least trying.

Microcampaigns as experiments

I’ve talked to a lot of founders who wanted to experiment with cold emails–but want to start with 10,000 contacts in their first campaign. There are good reasons for wanting to go big: to make sure the channel scales, to make sure they’re in a big enough market. Perhaps the agency they hired is only geared for larger scale outreach. But all of those good reasons do not make it a good idea.

The smart thing to do is to start small, with a microcampaign–or several. For example, testing messaging to different verticals. Sure, you’re horizontal platform could work for any industry. But which industries open your emails? Which take meetings? By using microcampaigns you can break down each industry into what works and what doesn’t, and make a better decision about where to scale up your efforts.

I did this years ago with the same product–dynamic pricing–and was shocked at the outcome. We tried 8 industries, 30-50 people in each. In one, retail fuel (gas stations), we couldn’t even find people to email. There was no job title that bought what we were selling. The results were awful. In another, hotels, we got a ton of responses but couldn’t get meetings–they were happy with an existing product, and it had thoroughly saturated the market.

My favorite industry, on-airport parking, opened all of our emails. I’m not kidding–like 80%+. They all wanted to meet. In 30 days we met with JFK, SFO, PSP, an airport somewhere in Wisconsin, and 20+ other significant airports. Pretty much anyone and everyone we asked. But it turned out they needed 2+ years to buy what we were selling, and all of them wanted us to fly there and see their parking lot–as if that mattered to our AI algorithms. Who has time for that? We needed to raise in 6 months!

By testing different strategies you can quickly figure out what works well and what doesn’t–without having to build out the scale and process required for a bigger campaign. Ideally individual AEs can run microcampaigns on their own.

How to run microcampaigns

Running microcampaigns is easy–it just takes a little thought. Start with a hypothesis. Who are your best customers? If you don’t have any yet, be thoughtful about who they might be. Why would they buy now? Who are the actual buyers–which people are involved? Could you try, for example, reaching out to people who influence other people, doing an indirect campaign. The doormat thing is kitchy and maybe it kinda worked. There are probably more on-brand, better ways to get attention.

Then, do the research to find the specific people. You can use ChatGPT or your favorite AI research tool to scour the internet. You can also just use Apollo.io or your favorite lead research tool to assemble a targeted list. You could get crazy and do both–use the AI to research, then Apollo to enrich with emails and LinkedIn profiles. Use fresh research; never buy a list, however tempting it sounds or how great their sample is. All bought lists suck. All of them.

Then, reach out. It can be as simple as writing 10 emails a day. Gmail has a built-in template tool. Or you can automate it, with a tool like skyp.ai. Remember that you have to follow up 3-5x to get a reply, because that is the state of the world: people ignore emails until they’ve seen it a few times. As much as I want to sell some software, you can absolutely do this by hand. One founder I spoke with recently sent 100 people emails with 2 follow ups entirely one-at-a-time through Hubspot. No campaign, no automation, just manually cut/paste/type. I honor the effort but have questions about whether this was the best use of her time.

With vibecoding, I find myself wanting to automate everything. All the time. While great, it can be a distraction. For microcampaigns go with easy. If writing 100 emails sounds like too many, just cut your list down to 50 or 25 people. You’ll still learn something.

The big caveat

Microcampaigns are unlikely to make your company a $1bn behemoth on their own. Instead, they’ll help you get off the ground, execute key experiments, and find that initial success that leads you to becoming the $1bn behemoth.

Because you bring the insights to the scaled campaigns, not the other way around. Whether you use skyp.ai, some other off the shelf tool, or even vibecode your own workflow, running tight, small campaigns will enable you to find that big, automated play that unlocks growth.