The Newsletter That Crashed the Website and Sold Out Everything

The website was down for 30 minutes. Customer service was overwhelmed. The client was terrified. And at the end of the day, every single item had sold out - the highest hourly revenue in the brand's history, with 70% of that day's sales attributed to a single email.

At our February INSIDE CRM Meetup in Berlin, Artur Suligowski walked us through how this happened - and why, despite the disaster in the middle, it turned out to be one of the clearest proofs of what a well-built newsletter program can actually do. Artur is the CMO at Cloudwise, a data-driven marketing and analytics agency based in Poland with a focus on e-commerce brands. He came over from Warsaw for the evening, which immediately told us this was a talk worth paying attention to.

The brand that doesn't do sales

Fluff is a sustainable Polish fashion brand - natural down, certified ethical sourcing, manufactured entirely in the EU. They were founded in 2016. When Cloudwise took them on as a client, they had fewer than 5,000 contacts in their database.

The brand also doesn't participate in Black Friday. They don't push aggressive discounting. Sales are extremely limited - by design - because the brand identity is built on conscious purchasing decisions. That's a genuine challenge for CRM. If you can't communicate price drops and rebates, the standard newsletter playbook for fashion (sale alert, promo code, urgency) doesn't apply.

Cloudwise's answer was to make the newsletter program the brand's actual CRM engine, using GetResponse, their marketing automation platform. And before they could run good campaigns, they needed a database worth mailing. Their newsletter conversion rate at the start was 6.6% - below the 15-20% typical for fashion brands they work with. Low enough that scaling campaigns efficiently required getting the right people into the list first.

Lead magnets that cost less than a coffee

The approach Cloudwise developed for Fluff came from an earlier experiment during the pandemic. For another fashion brand - By the Moon, targeting young mothers - they created a printable coloring book as a lead magnet when advertising costs were spiking across the industry and CPLs were becoming brutal. It worked: cost per lead dropped to under 20 euro cents across Europe and the US.

For Fluff, they adapted the approach. First a puzzle (CPL: 36 cents), then a trend book when they needed a refresh (CPL: 1.40 euros - not as attractive, but still manageable). The critical part wasn't just the creative - it was the distribution format. Cloudwise used native lead forms on Meta and later Google Ads. Compared to website pop-ups, native forms converted nine times faster. When someone is already logged into their social account, the email field pre-fills automatically. The friction is lower. The signal is cleaner.

Double opt-in rates ran at 92% in recent months. That number sounds great until you see what happened next: nearly 50% of users who completed double opt-in never opened the first email. They left their address to get the lead magnet and then disappeared. This is an important thing to track in any lead magnet campaign - the list grows, but you're building two different populations. Cloudwise dealt with this through a simple resend mechanic: if the first confirmation email wasn't opened, send it again. But the deeper issue is that growth metrics alone don't tell you how engaged your list actually is.

One useful aside from Artur: their ROAS on lead magnet campaigns this year has been running between 33 and 55 - well above their typical 20-something average for standard campaigns. The lead magnet campaigns were growing the database and generating sales at the same time. That's the argument for running them even when the creative costs more than you'd like.

51,000 people showed up to a secret sale

Once the database had grown significantly, Cloudwise convinced Fluff to try something they'd never done publicly: a secret sale. Only subscribers would know about it. Up to 30% off. No announcements anywhere else. No paid amplification. Just the newsletter.

The average basket for Fluff reflects a premium price point. These aren't impulse buys. Customers had been asking for a sale for a while. The brand was ready. Cloudwise sent the email.

In the first hour, 51,000 users hit the site. The WooCommerce store - which Fluff hadn't thought to have the hosting team upgrade in advance - couldn't handle it. The server went down for 30 minutes. Customer service was flooded with messages from frustrated shoppers who'd tried to pay and couldn't get through checkout.

Artur and the team were, in his words, terrified. They genuinely didn't know if Fluff would fire them.

They didn't. Because despite the crash, the orders kept coming in somehow - some users managed to complete purchases during the outage. By end of day: highest ever hourly sales in brand history. Newsletter attributed to 70% of daily revenue. The monthly newsletter revenue share jumped from its baseline of 6.6% to 22% for that month. Everything sold out.

Artur said the key lesson isn't "run secret sales" - it's "don't panic when something goes wrong." Not every disaster is as bad as it looks in the moment. If you stay calm and the underlying campaign is solid, the results might still surprise you. He also noted what they'd do differently: communicate the sale in advance to give people a window, give the server team more notice about expected traffic volume, and don't assume ecommerce operators have stress-tested their infrastructure just because you asked them to.

What you can take from this

The Fluff story is worth thinking about in two parts. The first is tactical: lead magnets built on genuinely useful content, distributed through native forms, can grow your database at a cost per lead that's hard to match with other acquisition methods. The second is strategic: if you build a newsletter program around a brand's actual identity instead of forcing the standard playbook onto it, the list you grow will behave differently. People on the Fluff list weren't there because they'd bought once. They were there because they trusted the brand - and when the brand offered something just for them, they showed up in numbers no one expected.

The server part is fixable. That kind of audience response is not.

You can find Artur on LinkedIn. If you want to hear these kinds of case studies in person - and ask the questions that don't make it into articles - insidecrm.io is where it happens.

Artur Suligowski is the CMO at Cloudwise, a data-driven marketing and analytics consultancy based in Poland specializing in e-commerce growth, CRM, and paid media. He has 10 years of experience in e-commerce and CRM across global companies and startups, including a stint at Zalando in Berlin.

Based on a talk by Jessica Jantzen  •  January 2026

“Everything, everywhere, all at once.”  That’s Scott Brinker’s movie-title diagnosis for how marketers feel about AI right now - and it’s hard to argue with. New tools every week, agents doing things that weren’t possible six months ago, and a nagging feeling that everyone else is further ahead.

But here’s what the data actually shows. Scott Brinker and Frans Riemersma’s 127-page MarTech 2026 report - one of the most thorough annual studies of the landscape - finds that most companies are not ripping out their stacks. Most are not running fully integrated AI agents. The majority are somewhere between “testing a pilot” and “using embedded features.” If that sounds like your team, you are in the majority, not behind.

Your stack isn’t going anywhere

The CRM, the campaign orchestration layer, the delivery infrastructure - all of it remains the critical data foundation. AI layers on top. What’s changing is not the foundation but what you can do with it.

Brinker’s research shows how adoption actually breaks down across practitioners:

• 62% use AI embedded in existing tools (HubSpot, Klaviyo, etc.)

• 49% use no-code/low-code solutions like Zapier, Make, or Workato to extend what they have

• 29% have internal development resources on AI topics

• ~10% are not yet using AI in any meaningful form

The typical marketer uses 3.5 tools. Hybrid stacks - a mix of embedded AI plus a no-code connector - are the norm. Highly customised, bespoke AI infrastructure is still the exception.

Where most companies actually sit

83% of companies are still in the “limited production to planning” zone. Only a small minority have AI agents fully integrated into core workflows. B2B companies are measurably further ahead than B2C - if you’re in retail, DTC, or a consumer brand, you have more runway than you might think.

The top use cases being deployed right now

• Content production -- copy, subject lines, A/B variants

• AI-optimised publishing and send-time optimisation

• Customer service chatbots for first-touch support

• Audience discovery and behavioural segmentation

• Competitive analysis and sales prospect research

Use this as a checklist: where are you already active? Where is there untapped potential?

The shifts you need to be tracking

Search is changing under your feet

McKinsey data shows around 50% of customers already use AI-powered search. That puts 20–50% of your traditional organic search traffic at risk. By 2028, consumer spend flowing through AI search is projected to shift substantially. This is not a future problem - it’s a now problem.

Email is getting more complicated

AI filtering is already routing messages differently. Preview text is being overwritten or hidden before your subscriber ever sees it. AI summaries are replacing the actual message. Tools like Superhuman are orchestrating inboxes and reducing how many emails people actually read.

The practical implication: you have less control over how emails appear than you did two years ago. The durable investment is trust and relevance - building communications your subscribers actively want, not ones that game the algorithm. One observation from the Berlin room: email could gradually shift from an “owned” channel to something closer to “paid” as deliverability and rendering control erodes. Channel diversification is becoming a necessity.

One thing worth learning about: MCP

Model Context Protocol is an emerging standard for how AI models connect to external tools - and it matters because it changes what “using AI” actually means. Right now, an LLM can write an email but you have to go paste it into Gmail yourself. MCP closes that gap: the model can pull live CRM data and take action - create the draft, update the record, trigger the send.

Several platforms already support it, including Klaviyo. You don’t need to run your own server to experiment - start with a vendor demo or managed solution. The video session includes a live MCP demo that makes this concrete.

The short version

AI agents are layering onto existing stacks, not replacing them. Your CRM is still the north star. The practitioners building good habits now - small instrumented pilots, solid governance, channel diversification - will be in the strongest position in two years.

Resources

• Scott Brinker & Frans Riemersma: MarTech for 2026 (primary research source)

• MCP explained - 9x (Berlin-based AI education, free weekly webinars)

• Tools referenced: HubSpot, Klaviyo, Superhuman, Zapier, Make, Workato

Presented at INSIDE CRM Meetup in Berlin, January 2026.

This article is designed specifically for CRM professionals and marketers who want practical, implementable ways to use AI across campaign planning, QA, personalization, and segmentation. By the end, you'll understand how to add AI to campaign workflows, where it helps most, what to watch out for, and simple starting patterns you can try this week.

Panelist introductions

• Audrey Mann: AI consultant with a background in CRM and seven years in marketing. She recently transitioned from running CRM programs to focusing on AI and automation adoption across organizations, helping teams implement content automation, campaign planning workflows, and QA checks.

• Felix Böhm: CRM specialist at a dedicated CRM agency with around eight years' experience. He focuses deeply on CRM strategy and implementation, and is currently building tools to help clients without ML teams adopt model-driven personalization and automation.

Watch it in action

🎥 Watch the complete panel with Q&A (paid members):

Resources

• Make (Make.com) - AI agents + automation

• Custom GPTs / GPT Store - reusable copywriting assistants

• Customer.io - AI-friendly segmentation and UX

Key takeaways

1. AI is most immediately useful in campaign content and QA workflows - copy scaffolding and automated checks save time.

2. AI agents (Make, automation platforms) let you run automated decision chains 24/7, but they require careful prompt design and iterative tuning.

3. Out-of-the-box AI personalization/decisioning can be a black box - prefer solutions that provide explainability or an audit trail.

4. Practical AI segmentation works best when combined with manual review and human-in-the-loop controls.

5. Small teams benefit most from templates, stored high-performing content, and lightweight custom GPTs.

Understanding how AI fits into CRM workflows

Most explanations of AI in CRM are either too technical or too basic. Let's be honest: you don't need a PhD to get useful results, but you do need to understand where AI provides value and where it can fail.

Think of AI in CRM as two complementary capabilities:

• Generative and assistive models for content (copy, subject lines, product recommendations).

• Automated decisioning and orchestration (AI agents, model-based routing, segmentation suggestions).

The practical sweet spot is when AI speeds up repeatable creative tasks and flags issues that would otherwise need manual QA.

Level 1 - Quick wins (what to build first)

1. Email QA automation

• Problem: Mistmatches between source copy and rendered email across markets and languages.

• How it works: Send test email to a staging inbox, trigger an automation that extracts rendered content, then have an LLM compare rendered output to source files and checklist rules (sender, links, asset presence, localized terms).

• Why it helps: catches broken links, missing assets, wrong sender names and simple localization mishaps before sending.

• Pro tip: Keep the QA checklist explicit in the prompt (e.g., "Check sender name, subject line tokens, link domains, alt text for images").

2. Copy scaffolding with a data-store of best performers

• Problem: Content writers get sparse briefs and re-write from scratch.

• How it works: Maintain a database of best-performing email templates and blocks. When a campaign manager fills a briefing form (campaign type, audience, tone), an automation queries the store and generates a draft tailored to those inputs.

• Why it helps: speeds up drafting and preserves what worked before.

• Pro tip: Surface the historical performance metrics alongside the suggested copy so writers can judge relevance.

3. Lightweight Custom GPT for repeat copy tasks

• Problem: Repetitive email text and microcopy waste time.

• How it works: Build a Custom GPT tuned for your brand tone + campaign templates. Use public GPT store prompts as inspiration to learn instruction structure.

• Why it helps: gives consistent, brand-aligned outputs and can be iterated quickly.

• Pro tip: Inspect popular GPTs in the store to reverse-engineer instruction sets and avoid starting from zero.

Level 2 - Automation + AI agents

AI agents let you connect event/data inputs, chain prompts, and trigger downstream automations (CRM updates, sends). They are essentially an automated orchestration layer that runs without manual prompts.

• Components: system prompt, helper prompts, files/data sources, scenario/automation modules.

• Workflow pattern: webhook/event → AI agent decides scenario → agent calls automation (e.g., update profile / trigger campaign) → automation returns output to agent → agent decides next step.

Practical notes:

• Use agents when you need 24/7 decisioning on streaming events (site behavior, incoming tickets).

• Hallucination is real: make sure outputs are validated against rules or human review before major actions.

• Iteration matters: building an agent is often about prompt engineering + workflow tuning.

Pro tip: Treat an AI agent like a new team member-log every decision, expose its reasoning, and give managers an easy audit view.

Level 3 - Personalization and AI decisioning

AI-powered product recommendations and channel decisioning can increase relevance, but many vendor implementations are black boxes.

• Use-case: CDP + AI decisioning to route users to the best channel (email, SMS, ads).

• Pitfall: Stakeholders want to know why decisions were made. If the AI is a black box, business will lose trust.

• Practical approach: prefer tools that offer an "AI analysis" or explanation layer so you can trace the features and rules that led to a decision.

Example pattern:

• Run AI decisioning to propose channel choices

• Save proposed decisions and model features to a review dashboard

• Run an A/B or holdout test to validate impact before full rollout

Pro tip: Always pair AI decisioning with a small, explainable sample check for management approval.

Segmentation with AI - realistic expectations

Some native vendor features (e.g., Dynamics Copilot) may produce poor segmentation if data fields are messy or prompts are weak. Better results come from platforms that rethink the UX and let you refine AI-suggested cohorts.

Practical recipe:

1. Clean and document the key fields the model can use (recency, frequency, monetary, product categories, engagement signals).

2. Ask the AI for a suggested segment using short, actionable prompts ("Find users likely to purchase product X in next 30 days")

3. Convert AI suggestion into rule-based filters you can inspect and tweak

4. Run a validation cohort and check actual outcomes

Pro tip: Don’t accept the AI’s cohort as final-use it to accelerate hypothesis formation and then apply manual safeguards.

Troubleshooting and common pitfalls

• Hallucination: Add specific verification steps and ground prompts to product IDs, canonical link domains, and approved phrasing.

• Black-box models: Prefer vendors that provide model-level explanations or a separate "analysis" view for audit.

• Over-automation: Keep humans in the loop for brand-sensitive or legally regulated communications.

• Prompt brittleness: Store successful prompts and parameter settings as templates so the team can reuse them.

Quick implementation checklist (first 30 days)

1. Map your campaign production process (ideation → copy → design → QA → send). Identify repetitive steps.

2. Start an email QA automation pilot: build a checklist and run LLM comparisons on test sends.

3. Create a small data-store of 10 best-performing email blocks and wire it to a drafting prompt.

4. Build a Custom GPT for basic email copy and test with 5 real briefs.

5. Log all AI agent decisions and create a simple audit dashboard for stakeholders.

Summary

AI in CRM is most useful when it speeds up repeatable creative work (copy scaffolding) and automates routine checks (QA, simple segmentation suggestions). The big opportunities are real, but success depends on careful orchestration: clear prompts, human-in-the-loop validation, explainability, and iterative tuning.

This comprehensive guide will walk you through how to collaborate with engineering and data teams so your CRM requests get built, tested, and prioritized without endless back-and-forth.

I'll assume you're a CRM practitioner who wants practical templates and a lightweight toolkit to translate business outcomes into the blueprints technical teams actually use. By the end, you'll know how to write requests they can estimate, measure, and ship and how to get better prioritization over time.

Watch it in action

🎥 Watch the complete presentation with Q&A (paid members):

Key takeaways

• Engineering and data teams think in blueprints: clear, precise requirements. Leave as little room for guessing as possible.

• Start every request with the business problem (X), not the solution (Y). Be open to alternative solutions.

• Structure requests as user stories + explicit data requirements + acceptance criteria.

• Use an impact hierarchy to help prioritize: revenue impact > user behavior > operational efficiency.

• Close the loop: share results and incrementality back with technical teams to build trust and future buy-in.

Understanding how technical teams work (quick primer)

• Engineering: goal is to build and maintain systems. They plan on quarterly or longer cycles, prefer precise requirements ("blueprints"), and deliver code, infra, and features.

• Data teams: goal is to build pipelines, transform data, and produce clean datasets and insights. Planning is slightly more flexible than engineering, but they also need clear specs on what data, where, and when.

Both teams measure success in output-oriented ways (deployments, pipelines, dashboards) and in business terms (adoption, conversion, system health). Your job as CRM is to translate outcome-focused asks into outputs they can implement.

Transition: So how do you make that translation? Start with the structure below.

Step 1: Start with business context (the most important part)

Always begin a request by answering: why are we doing this? Provide as much context as possible:

• Business goal: what outcome are you trying to change? (e.g., grow retail purchases among restaurant-engaged users)

• Success criteria: concrete target and timeframe (e.g., 50% of targeted users make a first retail purchase within 60 days)

• Current baseline: if known, state the current conversion or metric you want to improve

Why this matters: engineering and data often cannot infer the intended business impact. If you provide the why, they can propose better technical alternatives and prioritize appropriately.

Pro tip: Use this template sentence: "We want to [business outcome] because [business reason]. Success looks like [metric + target + timeframe]."

Step 2: Break the ask into user stories (who, what, why)

User stories remove ambiguity. Write requests as small, testable stories - not one huge vague request.

User story template:

• As a [persona], I want to [capability] so that [business outcome].

Example:

• As a CRM manager, I want a daily dataset of "engaged_restaurant_users" exported to the analytics warehouse, so that we can run a cross-retail automation targeting those users and measure first-retail purchase conversion.

Why user stories help:

• They frame the recipient (who), the deliverable (what), and the purpose (why).

• Engineers and data folks can split stories into tasks and estimate work more accurately.

• A great tool to break down a problem into smaller pieces and milestones.

• It helps scoping and prioritizing.

Step 3: Be explicit about data requirements

Data requirements are the most common source of rework. Spell this out:

• Precise segment definition: list the exact attributes and logic (e.g., engaged_restaurant_user = user who visited a restaurant page and performed action X at least twice in 30 days). Include references to an existing product or analytics definition if one exists.

• Frequency & freshness: daily / hourly / real-time? (be conservative — hourly is more expensive than daily)

• Destination & format: where should the data land? (warehouse table name, API endpoint) CSV/JSON schema, partitioning

• Volume & constraints: expected daily rows or unique users

• Integration points: which downstream system consumes it (marketing automation tool, campaign engine)

• Tests & validation rules: simple checks (row counts, null rates, sample records) and who signs off

Example data spec snippet:

• Segment name: engaged_restaurant_users

• Logic: users with >=2 "restaurant_checkout" events in last 30 days AND last interaction < 28 days

• Output: BigQuery table analytics.crm.engaged_restaurant_users (daily partitioned)

• Freshness: daily at 02:00 UTC

• Format: user_id, email_hash, last_activity, segment_start, attributes JSON

• Validation: row count >= X; sample of 100 records reviewed by CRM product owner

Pro tip: Add links to the data dictionary or screenshots of the analytics UI to remove ambiguity.

Step 4: Define acceptance criteria and "definition of done"

Acceptance criteria protect both sides. They state exactly how you and the technical team will know the work is complete.

Acceptance checklist (examples):

• Data exported to destination (table name + location) on agreed schedule

• Schema matches spec (field names and types)

• No more than Y% nulls in critical fields

• A sample of N records matches expected values

• End-to-end test campaign can be built using the dataset

• CRM owner confirms and signs off

Also state who is responsible for testing, and who closes the request.

Avoid the XY problem: start with the real problem

Engineers use the XY problem to describe when someone asks for a solution (Y) without explaining the real problem (X). The trap: Y may not solve X.

How to avoid it:

• Lead with the problem: "We are seeing a 10% conversion gap between A and B and need a repeatable way to identify users who are likely to convert."

• Follow with optional: "We think solution Y (e.g., add this segment) will help, but we are open to alternatives."

This opens the door for technical teams to propose better or cheaper solutions.

Impact hierarchy: how to get prioritized

When everything competes for engineering and data resources, impact framing matters:

• Revenue impact (highest priority): expected change in purchases, LTV, or revenue

• Pro-tip: choose a business metric that is known across the organization (e.g. incremental GMV)

• User behavior: engagement, activation, retention metrics

• Operational efficiency: hours saved, process automation

If possible, translate an operational ask into business terms like measure first in FTE and estimate what uplift in business metrics you could get from these FTE saved (e.g., automation that saves 40 hours/month could reduce time-to-launch and increase campaign velocity, enabling revenue-driving experiments). But be realistic: purely internal tools often sit lower in the backlog.

Pro tip: If your ask is operational-only, schedule it in advance and pair it with a small product/user-facing improvement to increase prioritization.

Collaboration patterns by company stage

Seed / very small teams:

• Communication is informal. Slack and quick meetings work. Documentation can be light.

• Keep requests small and run frequent alignment sessions.

• Business metrics might not be so important here as you are first looking for product-market fit. Everyone is aligned with the product vision/strategy.

Scale-up:

• More squads and specialization. Roadmaps are more rigid. Start formalizing user stories, data specs, and impact metrics.

• Invest in a living data dictionary and a lightweight ticketing process.

Enterprise:

• Formal planning cycles, SLAs, and service-level agreements. Requests need clear SLAs (e.g., time-to-deliver, support windows).

• Stop relying on Slack as the source of truth. Use documented templates and formal handovers.

Remember: what worked at 10 people will break at 100. Evolve processes as you scale.

Example: end-to-end request (short)

Business context: We want to increase retail purchases among users who frequently use restaurant features. Success: 50% of targeted cohort make a first retail purchase within 60 days.

User story: As a CRM manager, I want a daily segment of "engaged_restaurant_users" exported to the marketing DB so I can run a cross-retail automation.

Data spec: [see template above — include segment logic, destination, freshness, schema]

Acceptance: Table available daily; schema matches spec; one sample campaign successfully built and validated; CRM signs off.

Impact: Expected to add +X revenue per month by converting Y% of targeted users (estimate + rationale).

Common pitfalls and troubleshooting

• Vague language: "engaged users" or "recent activity" without definitions => rework.

• Last-minute deadlines: engineering plans in cycles. If you need something in 2 weeks, expect pushback.

• Not closing the loop: failing to share results makes future prioritization harder.

• Ignoring the data side: the dataset often takes longer than the campaign logic — plan accordingly.

Pro tips (sprinkled practical advice)

• Write small, testable stories. Smaller stories = faster delivery.

• Add business impact early in the ticket title (e.g., "[Revenue] 50% first purchase target") so it shows up in prioritization.

• Keep a rolling backlog of operational asks that you can trade into planning cycles.

• Use one-pagers with screenshots for complex product states - a picture removes ambiguity.

• When in doubt, schedule a 30-minute discovery call with an engineer or data engineer to align on scope before you file a ticket.

Summary

Working well with engineering and data teams is a skill you can learn. Focus on:

• Starting with the business problem, not a presumed solution

• Writing clear user stories and precise data specifications

• Defining acceptance criteria and who signs off

• Framing impact using the impact hierarchy

• Closing the loop and sharing results to build trust

Do these reliably and you'll reduce friction, speed up delivery, and get better prioritization.

About the presenter

Oscar Palomino is a Tech CRM Manager at Wolt, where he bridges the gap between CRM strategy and technical execution. He's worked across company stages from seed startups to scale-ups, and specializes in helping CRM teams translate business outcomes into technical requirements that engineering teams can actually execute on.

I asked our community about their biggest email deliverability questions. The overwhelming response? "How can deliverability problems be proactively avoided?" and "How can email deliverability be monitored?".

This post answers both. If you've been relying on your ESP's delivery rate as a success metric, you're missing the bigger picture. You'll discover why acceptance by an ISP doesn't mean inbox placement, why sender domain reputation now matters more than IP reputation, and how the right monitoring approach protects your ROI before problems escalate.

Thanks to Frank Rix from moin-inbox.de who is sharing his expertise of over 20 years in technical email marketing. Connect with him on LinkedIn.

In email marketing, all the clever copy and beautiful design mean nothing if your emails sink straight into spam. Deliverability is what keeps your campaigns afloat. And like any good ship, you need to keep an eye on every instrument to avoid drifting off course. That’s what monitoring is for.

The Basics of Deliverability Monitoring

Deliverability monitoring means knowing what really happens after you hit send. Do your emails reach the inbox or vanish into the spam abyss? A high delivery rate might look impressive, but it only tells you that the mailbox provider accepted your message, not that it reached the inbox.

Here’s what you need to keep an eye on:

• Inbox Placement – Where your messages actually land.

• Bounces – Hard bounces hint at bad data. Soft bounces mean temporary issues.

• Blocklists – End up on one, and your reach drops like an anchor.

• Authentication – SPF, DKIM and DMARC still separate the pros from the spammers.

• Engagement – Opens, clicks and replies show ISPs that your emails are wanted.

• Spam Complaints & Unsubscribes – The early warning lights on your reputation dashboard.

• Don’t just check once in a while. Trends over time are what tell the real story — especially when comparing mailbox providers.

Reputation Tracking in 2025

Reputation is still the make-or-break factor for inbox placement. But Google changed the game: the new Postmaster Tools no longer show those easy “domain” and “IP reputation” dashboards. Instead, you get compliance data like spam complaint rates, authentication status and encryption checks.

So if you want a reputation score today, you’ll have to build it yourself. Pull in metrics from Postmaster, Microsoft SNDS, blocklists and engagement data, then model your own view of sender health.

Professional deliverability tools do exactly that. They combine all those scattered signals and turn them into one clear reputation index — so you don’t need ten tabs open just to figure out what’s wrong.

Why Professional Monitoring Tools Matter

Sure, you could do all this manually. But unless you enjoy endless CSVs and guesswork, a professional monitoring tool will save your sanity. It gathers everything in one place, shows how Gmail differs from Outlook, and highlights the real trouble spots before they bite.

A professional monitoring tool does the heavy lifting for you. It consolidates data from different providers and perspectives, creating a holistic “big picture” of your deliverability. Instead of juggling isolated reports, you see trends, patterns, and correlations in one place. That makes it much easier to spot root causes, recognize risks early, and act on them before they snowball into bigger problems.

Here’s what they typically provide:

1. Reputation Tracking
   Good monitoring tools track your sender reputation across multiple sources. Since Google Postmaster no longer shows domain or IP scores, smart platforms model their own rating using signals like spam complaints, authentication success, and engagement. They also watch blocklists and alert you when your reputation drops.
   

2. Automatic Seed List Testing
   By sending your campaigns to a large network of test addresses across different ISPs and geographies, these tools show exactly where your emails land. This helps you identify provider-specific issues early on
   

3. Authentication Checks
   Automated verification ensures that SPF, DKIM, and DMARC are set up correctly and consistently – a must-have in today’s email ecosystem.
   

4. Engagement Metrics Beyond ESP Data
   While your ESP shows opens and clicks, professional monitoring solutions can help filter out bot activity and provide a clearer picture of real subscriber engagement.
   

5. Trend Analysis
   Over time, deliverability tools reveal patterns that might otherwise remain invisible – for example, gradual reputation erosion with a specific provider.
   

6. Aggregate Metrics
   Opens, clicks, open-rates, and click-rates viewed as time-series data show whether your reputation is improving or eroding.
   

7. Sending Patterns
   Sudden volume spikes or inconsistent frequency can trigger filters. Monitoring your cadence helps maintain trust with mailbox providers.
   

8. Mailbox-Provider Level Metrics 
   Performance can vary significantly between Gmail, Outlook, and regional providers. Comparing metrics per ISP allows you to pinpoint where issues originate.

The Business Impact

Deliverability isn’t just a technical metric. It’s a revenue factor. Every blocked or filtered email costs you money. Knowing your list weight, the share of your audience per mailbox provider, helps you measure how serious an issue really is.

If 30% of your subscribers use Gmail and Gmail starts filtering you, your entire campaign performance tanks. With proper monitoring, you can quantify the damage and the potential uplift once you fix it.

Final Thoughts

Email deliverability isn’t a one-time setup. It’s a living system that reacts to your content, your audience and how consistently you treat both. Monitoring keeps you alert, connected and ready to steer when the wind shifts.

Without it, you’re basically sailing blind through foggy waters, hoping to hit the right harbor. With the right monitoring setup, you’ve got radar, a clear horizon and full control of the wheel — ready to guide every email safely home to the inbox.

Hey there! You’ve probably noticed sometimes your emails don’t make it to the inbox. Maybe you’ve seen them land in the spam folder and wondered, “What the heck happened?” Well, you’re not alone. I’m not a full-time email deliverability expert, but I’ve been in the trenches for 12 years, sending millions of emails across different countries. I’ve even leaned on professional consultants to make sure things were done right. Now, I’m sharing my experience to help you avoid the mistakes I’ve seen and made and improve your own email deliverability.

Here’s what you’ll get from this guide:

Why Does Deliverability Matter?

Let’s cut to the chase: 1 in 6 legitimate marketing emails doesn’t make it to the inbox. That’s according to Validity’s report. And with an average revenue of $0.10 per email, those lost messages could be costing you a lot.

Imagine sending out 100,000 emails for a campaign—1 in 6 won’t reach its target. That’s a huge hit to your bottom line. So yeah, deliverability matters a lot.

Inbox placement is everything. Even if you think you’re doing well with a 98%+ deliverability rate, your messages might still not be reaching the inbox. Don’t trust those numbers at face value—dig deeper.

Why Email Deliverability Is Tougher Than Ever

We’re not living in 2005 anymore. Email has evolved, and staying on top of the game is harder than ever. Here’s why (Validity):

1. Increased Email Volume: Since COVID-19, global email volumes have doubled. That’s double the competition for inbox space.

2. Apple’s Mail Privacy Protection (MPP): This makes it harder to measure engagement through open rates, which can unintentionally harm deliverability.

3. Tough Economic Conditions: Let’s face it, people are more selective about the emails they open when money is tight.

4. Awareness of Deliverability Issues: More senders are catching on to these challenges, meaning the competition is getting smarter.

Be aware of evolving challenges. With everyone ramping up their email game, it’s not enough to just send messages—you need to send them smart.

How an Email Travels from One Server to Another

Sending an email is a lot like mailing a letter. Picture this: you write a letter, seal it in an envelope, and drop it off at the post office. From there, the post office figures out where your friend lives and routes your letter through a few stops along the way until it reaches the right mailbox.

Simple Mail Transfer Protocol (SMTP) works the same way for emails! When you hit "send," your email is passed off to a server (just like the post office) that knows how to get it to the recipient’s email server. Once it arrives, the email sits in their inbox, waiting for them to open it up and read.

So, in a nutshell, SMTP is the behind-the-scenes process that makes sure your email travels safely from your device to someone else’s inbox—just like the postal service gets your letter from point A to point B.

Now, I won’t dive too deep into the technical stuff, but the illustration shows a simplified version of what happens between the sending mail server and the receiving mail server.

https://postmarkapp.com/guides/everything-you-need-to-know-about-smtp

When Things Go Wrong

Of course, things can break along the way. If something goes wrong, you’ll get an SMTP error code, which lets you know whether it’s a temporary issue or a bigger problem. If you’re not sure where to find these error codes in your email service provider (ESP), don’t hesitate to reach out to their support team for help.

Bounced Emails: Why They Happen

You’re probably thinking, “Wait a second, is this where the bounces come into play?” Yep, you’ve nailed it! This is exactly when bounces happen, and they come in a few different flavors.

🟡 Soft bounces (4xx error codes)
These are the “not right now” kind of bounces. Soft bounces happen when there’s a temporary problem delivering your email. Maybe the recipient’s inbox is full, or the email server is down for a bit. The good news? The email server will usually try again later. The bad news? The message hasn’t been delivered yet, so it's hanging out in limbo for now.

🛑 Hard bounces (5xx error codes)
Hard bounces are a more permanent issue. They happen when an email is completely rejected because of something like an invalid email address, a non-existent domain, or just a plain wrong email. These emails aren’t going anywhere unless you fix the issue—there’s no second chance here.

🚫 Block bounces (550 error code)
This is when the recipient's email server slams the door in your face. It happens when your email gets rejected outright, often because of security reasons. Maybe your sender’s domain has been blacklisted, flagged as spam, or outright blocked. No matter the reason, the email won’t even make it past the front door.

Bounces—whether soft, hard, or block—only scratch the surface of what can go wrong in the email delivery process. SMTP error codes are the server’s way of giving you a status update on what went wrong, and they cover a lot more than just bounces. If you want to geek out over the full list of error codes, you can check them out here.

The Inbox Journey: Key Checks Before Your Email Arrives

So, we’ve covered how your email is sent, but before it lands in someone’s inbox, it has to pass through a series of important checks. Think of it like your email going through a quality control process—it needs to meet certain criteria to be delivered successfully.

Here are the five key checks every email faces:

1️⃣ Authentication
This is your email’s ID card. Without proper SPF, DKIM, and DMARC setup, email providers can’t confirm that the message is really from you, and that’s a big red flag. It’s the first line of defense for inboxes.

2️⃣ Reputation Check
Just like a reputation in real life, if your email sender reputation isn’t good, you’re going to run into trouble. If you’ve had a history of sending unwanted emails, or if people aren’t engaging with your content, your reputation takes a hit, and your deliverability will too.

3️⃣ Content Filtering
Email providers are like bouncers at a club—they’re checking to make sure your email doesn’t look suspicious. If your email has too many spammy words, all-caps, or excessive links, it could get flagged as junk before it even has a chance to make an impression.

4️⃣ User Engagement Metrics
How people interact with your emails makes a huge difference. Are they opening them, clicking links, or just ignoring them? Low engagement sends a message to email providers that your emails aren’t wanted, and that can affect future deliveries.

5️⃣ Inbox Placement Decision
This is the final checkpoint. After passing through all the filters, your email is either rewarded with a spot in the primary inbox or sent off to less desirable tabs like promotions or, worse, spam. All the previous checks lead up to this moment.

Let’s break these down further so you can make sure your emails are passing every check with flying colors.

1️⃣ Authentication: Making Sure Your Emails Are Legit

Back in the early 2000s, email systems were a bit like the wild west—anyone could pretend to be someone else. This led to a huge surge in email fraud and spam. People were sending fake emails left and right, tricking others into giving up sensitive info (this is where phishing attacks took off). So, email authentication was born to solve that problem.

Authentication helps ensure that emails really come from who they say they do, instead of someone pretending to be you.

Let’s break down the key players in this process:

〰️ SPF (Sender Policy Framework)
Think of SPF like a guest list. It checks if the email address sending the message is allowed to send emails on behalf of your domain. You basically tell the email system, “These are the addresses that are cool to send emails from my domain.” If an email comes from an unauthorized address, it’s going to get rejected.

〰️ DKIM (DomainKeys Identified Mail)
DKIM adds a digital signature to your email, like a stamp of approval that proves the email hasn’t been tampered with. It tells the recipient's server, “This email is legit and hasn’t been altered on its journey.”

〰️ DMARC (Domain-based Message Authentication, Reporting & Conformance) DMARC is like the referee—it enforces the rules. It uses both SPF and DKIM to decide whether an email should be delivered or blocked. Plus, it gives you reports on how your emails are doing, so you know if something's going wrong.

〰️ BIMI (Brand Indicators for Message Identification)
This is the fancy touch. BIMI allows your brand’s logo to show up next to your emails in the inbox—think of it like your brand’s badge of authenticity. But here’s the catch: BIMI only works if you’ve set up DMARC properly. It’s a nice way to build trust with your recipients and stand out in the inbox.

It’s all about making sure your emails are trustworthy and secure, so your recipients feel confident opening them.

How Can I Check My Authentication Setup?

At this point, you might be thinking, “Okay, but how do I know if my setup is good?” Well, let me introduce you to aboutmy.email. This tool will give you a health check of your email deliverability setup. It’s pretty cool because it pulls everything into one place and gives you all the nerdy details—though we won’t dive into all of that here.

Here’s what you need to do:

1. Head to aboutmy.email and generate a special email.

2. Send your newsletter to that email.

3. Wait for the results to pop up. You’ll get a summary of different checks and results, plus a sidebar with more categories to explore.

How Can I Improve my Authentication?

If your reports show error messages, don’t panic. There are plenty of free tools, like easydmarc, that can help you fix them. They’ll explain the different elements and guide you through what needs to be done.

I remember when I did my first email migration, and let me tell you, those abbreviations (SPF, DKIM, DMARC) were like a foreign language! But don’t worry—the tools will help you figure it out bit by bit.

Once your records are good to go, you’ll need to implement them. This means finding whoever manages your domain (for me, it’s GoDaddy, where I bought insidecrm.io). It’s usually a quick fix—about a minute job—but if tech puts you at the bottom of the list, maybe bribe them with a little chocolate to speed things up.

A Little Story on BIMI

I remember the old days when you needed a Google+ account just to see your logo in the Gmail inbox. It was a hassle, but hey, at least it was free. Fast forward to today, and now you’ve got to pay around $1,000 a year just to display your logo with BIMI. Not exactly a bargain.

I did some research, and honestly, I couldn’t find a solid study that proves BIMI makes a significant impact. You can’t really A/B test it either. Sure, you can insert the logo without the extra certifications, but no one’s really sure how much visibility that gets you. If you want to dig into whether it’s worth the investment, check out this great article over at All About Email.

2️⃣ My Reputation is Everything

Your sender reputation is like your email’s street cred—it’s how email service providers (ESPs) decide whether to trust your messages or toss them into the spam folder. Think of it as your email's "credit score." A good score opens doors, while a bad one slams them shut.

Your reputation is built on two main pillars: IP reputation and domain reputation. Let’s break it down:

🔢 IP Reputation
Every email you send out comes from an IP address. It might look something like this: 192.168.123.132. Your ESP assigns this to you, and it could be shared with other senders (shared IP) or exclusively yours (dedicated IP). This is the “home address” the server checks to determine if you’re a trustworthy sender.

Here’s the catch: it’s super easy to damage your IP reputation with just a few bad emails, but fixing it can take weeks or even months. So, protect it like gold!

🏡 Domain Reputation
Your domain, like mine—insidecrm.io—also needs to have a good reputation. The reason? You can swap out IPs and get fresh ones, but your domain stays tied to your sending behaviour. If your domain starts to look shady, you’ll have a tough time getting into inboxes, no matter how many new IPs you buy.

What Factors Impact My Reputation?

Here’s what can make or break your sender reputation:

〰️ Spam Complaints
When people click “Report as Spam,” it’s like dinging your credit score. The more complaints, the worse your reputation gets. Keep that number low to stay in good standing.

〰️ Spam Traps
These are sneaky email addresses set up by ISPs to catch spammers. They look like normal email accounts, but if you send to one, your domain could get blacklisted fast.

〰️ Bounce Rates
Bounces happen when emails don’t make it to their destination. Soft bounces are temporary (think full inboxes), but hard bounces are permanent (bad or non-existent email addresses). Too many of either, and your reputation takes a hit.

〰️ Engagement
This is all about how people interact with your emails—do they open, click, or reply? Strong engagement shows ISPs that your emails are wanted, which helps your reputation.

〰️ Sending Volume and Consistency
Sending too many emails too fast or having irregular sending patterns can raise red flags with ISPs. Consistency is key—build trust over time, don’t go overboard in one go.

〰️ Unsubscribe Rates
While not as damaging as spam complaints, high unsubscribe rates are still a warning sign. If lots of people are jumping ship, ISPs might take notice.

The Unsubscribe Dilemma

Let’s talk about unsubscribes for a second. A lot of email marketers make it hard for users to unsubscribe, thinking it’ll keep people on their list. But honestly? That’s a quick way to tank your sender reputation.

Here’s why: if someone doesn’t want your emails anymore, they’re going to find a way out. If they can’t easily unsubscribe, they might mark your emails as spam, or worse, just stop engaging altogether. Both of these actions hurt your reputation.

Google Postmaster: Your Reputation HQ

If you’re serious about keeping tabs on your sender reputation, you need to know about Google Postmaster. Since Gmail is one of the most popular inbox providers (about 50% of email addresses at many companies are Gmail accounts), it’s a good idea to monitor your reputation specifically with Google.

Google Postmaster is a free tool that lets you see your reputation with Google domains (like @gmail.com). It shows you data on authentication, spam rate, and unsubscribe behavior, so you can spot issues before they spiral out of control.

Here’s what you can expect:

• Compliance status: A checklist that tells you if your emails are meeting Google’s standards based on factors like authentication and spam rate.

• Spam rate: Shows how many of your emails are marked as spam.

For the nerds out there, you can dive deeper by visiting the “Old Postmaster Tools” section, where you can check your IP and domain reputation. The setup is simple, but it takes a few days for the data to start showing up, so don’t worry if it doesn’t appear right away. You might not need to check this every day, but it’s a lifesaver when your performance dips unexpectedly, and you need answers fast.

3️⃣ Content Essentials: Craft Emails That Land in the Inbox

Can I tell you exactly which content elements will determine whether your email lands in the inbox or the spam folder? Not 100%. But I’ve got a solid checklist of best practices that can give you a real edge. These tips will help improve your chances.

Here’s what you should keep in mind:

✅ Subject Line and Preheader Length
Keep your subject lines and preheaders within recommended character limits. Too long, and they’ll get cut off—especially on mobile. Make sure your message is clear and compelling in the space you have.

✅ Avoid Spammy Language
Certain words are like red flags to spam filters. Words like “FREE,” “WIN,” and using all caps can trigger those filters, sending your email straight to spam. Keep it conversational and professional.

✅ Keep Links Consistent
Make sure all your links come from the same domain. If your email contains links from multiple domains, it can raise suspicions and affect deliverability.

✅ Use Alt Text for Images
Every image in your email should have descriptive alt text. This not only helps with accessibility but also ensures that your message still gets across if images don’t load.

✅ Balance Image-to-Text Ratio
A big no-no is sending image-heavy emails with hardly any text. Keep a balance—too many images and not enough text can get your email flagged by spam filters.

✅ Watch Your Email Size
Keep your email under 102 KB. Why? Because Gmail cuts off emails that are too big, forcing users to click a “View Entire Message” link. That’s an extra step for your readers that you want to avoid.

✅ Design for All Devices
Your email needs to look great on every device, whether it’s a desktop or a phone. Following mobile-friendly best practices ensures your email gets read, not deleted.

✅ Keep Width in Check
Stick to a maximum width of 600px for desktop and 320px for mobile. Wider than that, and your email could end up looking wonky on different devices.

✅ Clean HTML Code
Messy or broken HTML can break your email layout, make it unresponsive, and even get flagged by spam filters. Double-check your code and keep it clean.

✅ Easy Unsubscribe
Make unsubscribing simple. Don’t hide the unsubscribe link. A difficult unsubscribe process can lead to spam complaints, which hurt your reputation and deliverability.

Following these best practices gives you a better shot at getting your email delivered and read. Sure, inbox placement is a complex game, but taking these steps gets you closer to winning it.

4️⃣ Engagement Rules: How Your Audience Impacts Your Email Reputation

When it comes to email marketing, engagement metrics are like your report card. They tell the world (and the ISPs) how your emails are performing. If your engagement is high, it’s a green light for your sender reputation. If it’s low? Well, that’s when deliverability can start to take a hit.

Let’s break it down into positive and negative engagement metrics, and how each one impacts your reputation.

🚀 Positive Engagement Metrics

➕ Open Rates
A high open rate shows that people actually want to read your emails. It’s like telling ISPs, “Hey, these emails are valued,” and that gives your sender reputation a solid boost.

➕ Click-Through Rates (CTR)
It’s great if people are opening your emails, but even better if they’re clicking through to your content. A strong CTR means your emails are not just getting opened—they’re engaging, which reflects positively on your reputation.

➕ Reply Rates
When recipients hit reply, it’s a big win. This kind of interaction shows that your emails are personal and valuable. Pro tip: Use a real, personalised “Reply-To” address to encourage more responses.

➕ Forwarding
If people are forwarding your emails, that’s gold. It’s a major signal to ISPs that your content is worth sharing, and that can only help your reputation.

➕ Time Spent Reading
The longer people spend reading your emails, the better. ISPs factor in how much time recipients are engaging with your content, so aim to create emails that hold attention.

🚨Negative Engagement Metrics

➖ Spam Complaints
If recipients are marking your emails as spam, that’s a major red flag. It’s like telling ISPs, “This email is unwanted,” and that can do serious damage to your reputation.

➖ Bounce Rates
A high bounce rate—especially hard bounces (invalid email addresses)—hurts your sender reputation. It signals that your email list might be outdated or full of inactive addresses.

➖ Low Engagement Rates
If your open rates and click-through rates are consistently low, ISPs start to get the hint that your emails aren’t resonating with your audience. This can negatively impact your deliverability over time.

➖ Deletion Without Reading
If recipients are constantly deleting your emails without opening them, it’s a bad sign. ISPs will take this as a signal that your content isn’t relevant, which can hurt your chances of getting into the inbox in the future.

5️⃣ Inbox Placement: Getting Into the Right Tab

Back in 2013, Gmail shook up the email world by introducing the Promotions tab. Marketers everywhere panicked, wondering if their emails would ever be seen again. And here we are, still asking, "Did I land in the inbox, or am I stuck in Promotions?"

Now, Gmail doesn’t reveal the full details behind its sorting algorithm, but we do know that it uses the content of your email and the sender address to decide where to place incoming messages.

Here’s a quick breakdown of Gmail’s inbox tabs:

• Primary tab: This is where the magic happens. It’s for personal emails and important notifications—basically, the emails people pay the most attention to.

• Social tab: For messages from social media platforms like Facebook, Twitter, and Instagram.

• Updates tab: Here’s where transactional emails like shipping notifications or changes to shared Google Docs land.

• Forums tab: This one’s for emails from discussion groups, forums, or mailing lists.

Why the Primary Tab Matters

When we talk about inbox placement, we’re really talking about landing in the Primary tab. That’s where you want to be. According to a survey by Zerobounce, 25% of people never check their Promotions tab. And as for the spam folder? Forget it—27% check it just a few times a year, and 15% never check it at all.

Here’s the thing: delivery rate (whether the email was accepted by the server) doesn’t tell you where it landed. For that, you need to track your inbox placement.

Tools for Inbox Placement

If you’re serious about figuring out where your emails are ending up, tools like Evernest (formerly 250OK) can help. They provide a seed list of email addresses for you to send emails to. These addresses report back whether the emails landed in the inbox, Promotions, or somewhere else. You can track this data on a dashboard to get an idea of your inbox placement rate.

But a word of advice: this step is really more useful for larger senders. There’s plenty you can improve on before you need to dive this deep.

What I Didn’t Cover

There’s a lot more to unpack when it comes to email deliverability. But for now, I’ve focused on understanding the basics and some quick fixes. Here are a few topics I didn’t touch (or only briefly mentioned):

• List hygiene & segmentation

• IP warming

• Spam traps

• Sending history

• Double opt-in best practices

• Unsubscribe management

• Blacklisting

• How to handle high bounce rates

• Feedback loops

If you like to see a Guide 2.0 covering these topics, leave a comment on LinkedIn.

Key Takeaways from the Guide

Now that we’ve covered the essentials, here are the key takeaways from everything we’ve talked about:

1. Deliverability matters: Landing in the inbox (especially the Primary tab) is crucial. Even a high delivery rate doesn’t guarantee visibility.

2. Reputation is everything: Your sender reputation, whether tied to your IP or domain, is the biggest factor in whether your emails make it to the inbox. Keep it strong by avoiding spam complaints and bounces, and by maintaining engagement.

3. Authentication is a must: Setting up SPF, DKIM, and DMARC ensures your emails are trusted by email providers. Without proper authentication, your emails might never make it past the gate.

4. Content matters: Use best practices like avoiding spammy language, keeping your subject lines concise, and balancing image-to-text ratios to improve your chances of landing in the inbox.

5. Engagement is key: Positive engagement metrics like high open rates and click-throughs boost your sender reputation, while negative metrics (spam complaints, high bounce rates) can hurt your deliverability.

6. Inbox placement tools: For larger senders, tools like Evernest can help track exactly where your emails are landing, but make sure you’ve mastered the basics before diving into this level of detail.

Enjoyed the read? Fuel my next piece with a coffee! (A small way to help me keep creating great content!)