In an endless sea of content, nothing quite matches the grandeur, spectacle, and humanity of live events. No other advertising experience can replicate the impact of delivering your brand’s message alongside nail-biting moments or even captivating points in history.

It’s no wonder that publishers, SSPs, and DSPs are all clamoring for the opportunity to activate programmatic advertising on live events.

I’ve had the opportunity to sit at the bleeding edge of programmatic live-event activation, and it gives me a clear view of what’s working well and the opportunities for the ecosystem to improve. One area ripe for improvement is creating a standard way for publishers to share live event information with their ad technology partners before an event occurs.

Programmatic advertising is built upon standardization. A set of predefined protocols that tell each participant exactly how to act and what to expect when requesting ads, providing bid responses, playing back creatives, and so on.

Until recently, there was no standard for communicating live event information before an event occurred — until the Forecasting API.

What is the Forecasting API?

The Forecasting API was born out of the LEAP initiative from the IAB Tech Lab and offers a standard specification for publishers to share live event schedules, expected peak streams, and other event metadata with their ad platform partners.

The Forecasting API is one of multiple current and planned LEAP projects aimed at standardizing live event digital advertising. But what problem does it solve?

Why do we need the Forecasting API?

Live events introduce scaling, pacing, and optimization challenges for publishers, ad servers, SSPs, and DSPs that they do not face in VOD environments.

These problems lead to poor ad experiences for users or a failure to meet campaign goals for advertisers and publishers.

If the ad platforms involved in a live event only loosely understand what volume of viewers to expect, then they cannot properly prepare for the avalanche of traffic that can melt servers and throw bidding optimizations out of whack.

Publishers can email or share spreadsheets of event schedules and expected concurrency or QPS (queries/bid requests per second) with their partners, but this is time-consuming, inconsistent across publishers, and requires manual intervention.

The forecasting API aims to standardize this process and enable publishers to share schedules, event information, and expected concurrency with all their partners in a scalable way.

Preparing for live events

Understanding expected concurrency helps both the supply and demand sides ensure readiness. Not properly preparing for live events can cause many issues. Massive numbers of users streaming the same event concurrently can lead to:

• Dropped requests

• Increased timeouts and ad errors

• Inconsistent delivery and pacing

These issues can lead to dreaded slate during broadcasts (“We’ll be right back!”), which means lost opportunity for publishers.

Enabling new DSP features

DSPs also want up-to-date scheduling information so they can build features to better support their clients' ambitions for live event buying. These features include new live event targeting experiences and more intelligent pacing.

If DSPs have an automated, accurate method for fetching scheduling information, they can craft a more elegant in-platform buying experience by piping that schedule information into the platform.

Instead of a scattered deal-by-deal setup across publishers, DSPs could instead create solutions to buy across an entire league. A DSP could, in theory, allow an advertiser to simply target “NFL” or “NBA” and then orchestrate targeting across multiple publisher deals in the background.

Additionally, by parsing scheduling information, expected viewership, and ad load, the DSP could more intelligently pace campaigns. Pacing presents a unique challenge when monetizing live events today, given unpredictable inventory availability and spiky bursts of traffic.

DSPs can better tune pacing algorithms not only to handle concurrency when it's known in advance, but also to spread out their clients' budgets throughout a season or even during special events like tournaments and playoffs.

How does the Forecasting API work?

Let’s dive into the specification and take a look at how this thing is supposed to work.

Someone is expected to host an endpoint that, when called, returns schedule and event information. Just like an SSP calls a DSP ORTB endpoint for a bid, a DSP can call a Forecasting API endpoint to obtain schedule and event information.

Who owns this endpoint and populates the scheduling information? It is somewhat flexible according to the spec.

Streamers could host it themselves or enlist their SSP to build it out. In any case, there needs to be a system in place for a publisher to populate schedule information along with event metadata, and host a service to return a response to authorized consumers of that information (most likely DSPs).

Here is an example response:

{
  "version": "1.0.0",
  "timestamp": 1751205600000,
  "upcomingevent": [
    {
      "id": "content_management_system_id_game1",
      "scheduledstart": 1751044800000,
      "scheduledend": 1751059200000,
      "flexibleend": 1,
      "eventstatus": 1,
      "content": {
        "id": "content_management_system_id_game1",
        "title": "TeamA vs TeamB - Series Name - Game 1",
        "series": "Series Name 2026",
        "season": "2026",
        "gtax": "9",
        "genres": "483",
        "language": "en",
        "producer": {
          "name": "Content Name"
        },
        "network": {
          "name": "Network Name"
        },
        "len": 10800
      },
      "streamsdata": [
        {
          "country": "USA",
          "expectedpeak": 2000000
        }
      ],
      "adinventoryconfig": {
        "supportedmtype": [
          2
        ],
        "totaladdurationsec": 1400,
        "expectedpodcount": 12,
        "unplanned": 1
      },
      "lastmodifieddate": 1750958400000
    }
  ]
}

Key fields worth noting

You’ll notice the response provides:

• Scheduled start and end times (scheduledstart + scheduledend in UNIX timestamp)

• Event title (title)

• Genre (genres 483 = American Football)

• Broadcasting network (network)

• Duration (len)

The event “ID” field is also important, as it ties the event in the forecasting API to a content ID in a bid request. This would give DSPs the capability to set up targeting in advance using scheduling information and target the specific event ID they see in the content.id field in a bid request.

However, the willingness of publishers to share content ID is up for debate. Also, if they did, it would almost certainly have to be a hashed identifier so it doesn’t reflect the internal value present in a publisher’s content management system.

StreamsData

"streamsdata": [
  {
    "country": "USA",
    "expectedpeak": 2000000
  }
]

The StreamsData object provides the crucial figure of peak estimated concurrent streams during the event. This is the magic number that allows ad systems to prepare for the high concurrency of a given event.

It’s worth noting that there is a separate Concurrent Streams API that could be used to validate this number in real-time.

AdInventoryConfig

"adinventoryconfig": {
  "supportedmtype": [2],
  "totaladdurationsec": 1400,
  "expectedpodcount": 12,
  "unplanned": 1
}

The AdInventoryConfig object provides insight into the ad opportunities available in a broadcast.

In this example:

• supportedmtype = 2 indicates video

• Total ad duration is 1,400 seconds (~23 minutes)

• 12 ad pods are expected (expectedpodcount)

• Additional unplanned pods may occur like injuries or stoppages (unplanned=1)

This information can help inform pacing decisions based on available supply.

This AdInventoryConfig can help inform pacing based on availability. DSPs could use this information to help spread out delivery across individual events or even entire seasons.

It could also help inform bidding behavior by providing a sense of inventory scarcity if they know exactly how much opportunity is available during an event. However, I will note that it will not be a straightforward exercise for DSPs to use this info for bidding and pacing, as DSPs won’t know how often direct IOs and competing DSPs fill these pods.

So the AdInventoryConfig and StreamsData objects alone won’t provide all the necessary information to optimize bidding and pacing. DSPs will need to synthesize Forecasting API information with past auction bidding and delivery data to unlock the new insights it could provide.

It’s now up to publishers and ad platforms to decide if the Forecasting API provides enough value to warrant adoption and integration into existing workflows and systems.

The API offers an automated, scalable path to replace a dreadfully manual current reality and could also provide additional insight into inventory availability, unlocking more advanced DSP targeting, planning, and bidding features.

If you want to read the full proposal or ask any questions, the Forecasting API is now open for public comment.

EVENTS

Marketecture Live III

7 Days Until Marketecture Live

The agenda is set, tickets are moving fast, and The Glasshouse is filling up.

Sponsored by AI Digital

Marketers are being pulled in more directions than ever before. Walled gardens are consolidating power, the media landscape is fragmenting at speed, and emerging technologies—from AI-driven buying to new identity solutions—are reshaping the rules faster than teams can rewrite their playbooks. In this environment, clinging to a single platform or closed ecosystem isn't just limiting—it's a strategic liability. The Open Garden Framework offers a way forward.

The Ad Tech Explained team developed this guide in partnership with AI Digital to unpack the Open Garden Framework—what it is, why it's essential in today's landscape, and how brands can put it into practice.

Sponsored by AI Digital

Let’s start with the basics: What is the Open Garden Framework?

AI Digital built the Open Garden Framework on a simple conviction: No single platform should own your strategy. It's a KPI-first approach to programmatic planning, activation, and measurement—one that treats the entire ecosystem as available inventory and selects the best combination of technology, data, and supply for each objective. The brand's goals lead. Everything else follows.

To be clear, the Open Garden is not a DSP, a buying platform, or a proprietary system. It's an operating principle that strips away the forced constraints of minimum spends, platform dependencies, and vendor bias. Every decision, from supply path optimization to audience strategy to cross-channel measurement, is made for one reason: because it's the best way to deliver against the business outcome.

In short, Open Garden is the modern philosophy of programmatic, free from the limitations of walled gardens and powered by the freedom to use the best-in-class technology, data, and inventory across the entire digital ecosystem.

What has changed within the marketing ecosystem that requires a new approach to “business as usual”?

For years, programmatic felt simpler because much of the inventory and audience access could be reached through a small set of platforms. That world is gone. Today, major platforms are closed ecosystems with their own rules and incentives, and cross-channel measurement is harder because data and signals are increasingly siloed or restricted.

At the same time, retail media has exploded into many networks with different logic and reporting. CTV, mobile apps, and the open web often require different buying paths. Ultimately, it can take five to seven tools just to reach consumers where they actually spend time.

In general: Control has consolidated, supply can become commoditized, signals can be unreliable, and intermediaries can skew the bidstream in their favor. The result is less visibility and governance across platforms, less flexibility, and rising operational overhead. Walled gardens still have value, but they can’t be the only answer anymore.

How does the Open Garden Framework solve modern marketing challenges?

Open Garden restores a critical advantage in a fragmented ecosystem: choice. Rather than letting a platform dictate strategy through lock-ins, minimums, or default settings, the approach ensures brand KPIs determine the tech stack and the plan.

Because it operates across channels, DSPs, retail networks, and CTV environments, it helps brands stay resilient when any single platform changes rules, restricts signals, or shifts incentives. It also enables AI-driven planning and optimization to work across the full media mix, not just inside one closed environment.

The practical effect is a simpler operating logic for complex times: Unify around the business goal, then select the best partners and paths to deliver it, with governance and measurement designed to compare performance across the ecosystem.

Sponsored by AI Digital

What are the benefits of the Open Garden Framework for brands and agencies?

The Open Garden Framework, when put into action, creates multiple tangible advantages:

• Cross-platform access: Reach all relevant channels, inventory sources, and data partners without political or commercial constraints.

• Freedom to choose the best: Select partners based on what’s best for the campaign, not what a single platform makes easiest.

• KPI-led decisioning: Start with the business objective and build everything else around it, not the reverse.

• Resilience in a fragmented landscape: Reduce dependence on any one ecosystem, protecting strategy from platform disruption.

• AI-accelerated strategy and execution: Apply AI across planning, supply curation, optimization, and analytics to automate complexity and spot better paths faster.

• Better economics: Reduce overhead and inefficiency, improving outcome per dollar while helping agencies avoid the heavy burden of stitching together fragmented programmatic.

Let’s talk about how AI Digital plans to put the Open Garden Framework into practice. What’s on the horizon?

AI Digital’s vision is an all-in-one AI marketing intelligence platform that puts the Open Garden Framework into motion by connecting research, planning, optimization, and reporting in one place, with no added platform cost. The right platform will be designed to give teams a unified workflow: understand the market and audience, build a cross-channel plan, activate and optimize, then prove impact.

Marketers need a solution that centralizes research and intelligence (audience segments, audience personas, inventory evaluation, competitive analysis). It should support strategic planning with AI-assisted media planning, forecasting, and cookieless targeting workflows. And it should tie in optimization and reporting, including campaign analytics, media mix modeling, brand lift, and path-to-conversion views so teams can move beyond vanity metrics and connect media decisions to real outcomes.

That’s what modern marketers are going to need to operate in today’s fragmented reality, and that’s what we’re building at AI Digital today.

About AI Digital

AI Digital is a AI-native media consultancy and innovation partner powered by an Open Garden Framework. Founded in 2018, AI Digital’s global team of 450 specialists connects clients to audiences through a unique blend of human expertise and advanced technology, delivering unbiased, AI-enhanced media solutions for brands and agencies. The company partners with all major advertising platforms to deliver measurable results aligned with clients’ business objectives and brand KPIs.

Sponsored by AI Digital

Sponsored by Triton Digital

Podcasts now reach more than half of the overall U.S. population every month. For advertisers, that means podcasting long ago stopped being an “add-on” channel. It’s a mainstream behavior with real scale.

The bigger story is who that reach includes. Podcast audiences over-index among younger adults and other “premium” segments, making the channel a strong fit for brand building and performance outcomes alike. And because podcasts now span audio and video consumption, advertisers can show up in moments that feel more like lean-in time than background noise.

The Ad Tech Explained team developed this explainer, in partnership with Triton Digital, to explore the current state of podcasting and what advertisers need to know about making the most of these high-value audiences. The insights below are informed by Triton Digital’s latest U.S. Podcast Report, which can be downloaded here.

Sponsored by Triton Digital

Let’s start with the basics. How is podcast consumption evolving as we move deeper into 2026?

At the top of the funnel, the channel continues to broaden. Triton’s data shows 53.6% monthly reach across the U.S., with especially high penetration among 18–34-year-olds (68.7%) and 35–54-year-olds (62.1%). Even 55+ audiences are participating meaningfully (35.0%), reinforcing that podcasting is no longer “just for the young.”

But “evolving” isn’t only about more people tuning in. It’s also about how they tune in: Today’s podcasting ecosystem includes traditional RSS listening, in-app streaming, and a growing habit of watching podcasts. That multi-format reality is what will matter most in 2026 planning: Podcasting is expanding, but it’s also fragmenting across formats, platforms, and genres.

How do podcast consumers compare to the overall U.S. population?

Compared to the general population, podcast listeners skew:

• Younger

• More educated

• More affluent

• More diverse

This is why podcasting keeps earning the “premium environment” label. It reliably aggregates audiences that many advertisers are actively trying to reach efficiently. If you’re using demo and behavioral signals to find high-value consumers, podcasting is increasingly a place where those signals over-deliver.

What do we know about new podcast consumers versus those who have been listening for years?

New listeners are helping reshape the audience profile. New listeners skew younger (18–34), while long-term listeners are most prominent in the 35–54 age range. Also, the gender gap continues to close among new listeners. 

More importantly for advertisers, new listeners also bring fresh commercial opportunities. Compared with more tenured listeners, past-year listeners show higher stated intent across multiple categories, including:

• Switching wireless providers (22% past-year vs. 15% among 5+ years)

• Online shopping (49% past-year vs. 38% among 5+ years)

• Quick-service restaurant visits (69% past-year vs. 60% among 5+ years)

That combination of audience growth and high-intent behaviors is a big reason podcasting is becoming more central to performance-focused plans, not just awareness buys.

Let’s talk modes of consumption. How is video factoring into the podcast consumption these days?

Video is real, growing, and meaningful, but it’s not a clean “podcasting is video-first now” story.

Triton’s data shows:

• 13% of monthly podcast consumers (18+) are listen-only

• 7% are watch-only

• 80% both listen and watch 

Even with the rise of video, audio remains the primary mode when you look at time spent: About 54.6% listening vs. 45.4% watching. In other words, podcasting is a blended ecosystem where audio still anchors consumption, while video expands entry points and boosts total engagement.

How do podcast consumers differ by method of consumption?

The audience isn’t evenly distributed across formats. Triton’s findings show that exclusive watching skews younger and male, while exclusive listening skews older, more female, and higher-income. 

Tenure also matters. Newer podcast consumers are more likely to lean into video, while more tenured listeners tilt more audio-first (for example, the share who are listen-only increases among 4+ year consumers, while watch-only is higher among past-year consumers). 

The podcast audience is not one monolith. If you buy only one format as an advertiser, you might be selecting for a meaningfully different audience profile than you intended. The best plans increasingly treat audio and video as complementary levers, not substitutes.

Sponsored by Triton Digital

There are many ways to tune into podcasts these days. Which platforms are being used most? And how do consumers vary across those platforms?

Among platforms, YouTube leads and is growing. Its share rose from 28.1% in 2022 to 37.7% in 2025 as the platform consumers use most often for podcasts, outpacing Spotify (24.6% in 2025) and Apple Podcasts (11.3% in 2025). 

But platform leadership depends on what you’re measuring. In new episode RSS download activity, Apple Podcasts still dominates (about 49.7% share), with Spotify next (18.1%), and the rest spread across players like iHeartRadio, Overcast, and others. 

Audience composition also varies: Spotify attracts a younger audience, YouTube skews more male, and Apple reaches higher-income users.

Let’s connect the dots. What does all of this mean for advertisers and how they’re building podcasts into their media plans?

Here are three practical takeaways:

1. Plan podcasting for scale and premium reach, not as a niche add-on. With 53.6% monthly reach, podcasting can carry meaningful weight in national plans, especially for brands prioritizing younger, educated, and affluent consumers. 

2. Buy format with intention, as audio vs. video is a genre-by-genre decision. Categories primarily consumed via audio include Science (58%), History (56%), Fiction (54%), Arts (51%), and True Crime (50%). Meanwhile, Music (34%), Sports (32%), Kids & Family (31%), Comedy (30%), and News (30%) skew more toward exclusive video consumption. Your “best” format depends on what people are doing in that content neighborhood. 

3. Use new listeners as a growth engine for performance. Past-year listeners show higher intent across categories like wireless switching, ecommerce, and QSR, creating a natural runway for advertisers looking to reach in-market consumers as the audience base expands. 

Finally, the ecosystem’s top players still matter for reach strategies. In Triton’s U.S. Podcast Ranker, iHeart Audience Network ranked No. 1 in average weekly downloads among sales networks. NPR News Now was the most downloaded podcast overall in 2025, and Bulwark Takes led new debuts.

About the 2025 U.S. Podcast Report

The 2025 U.S. Podcast Report combines insights from IAB Tech Lab–certified Triton Podcast Metrics, Triton Podcast Metrics Demos+, and surveys of 12,000+ monthly U.S. podcast listeners to deliver a detailed view of consumption behavior and audience opportunity across the ecosystem.

Whether you’re a creator, advertiser, or podcast fan, this report gives you the insights you need to stay ahead, spot opportunities, and make smarter decisions in an ever-evolving industry. Download today!

About Triton Digital

Triton Digital is the global leader in digital audio, podcasting, and broadcast radio technology. Trusted in over 80 countries, the company helps broadcasters, podcasters, and streaming services grow audiences, maximize revenue, and streamline operations.

Triton Digital’s industry-leading tools, Webcast Metrics® for streaming measurement and Podcast Metrics, one of the first IAB-certified podcast services, set the standard for data-driven success in online audio.

Learn more at TritonDigital.com.

Sponsored by Triton Digital

A series about building a real career in ad tech, navigating chaos, learning fast, and finding ways to grow without burning out.

People ask me fairly often how to start or advance a career in ad tech. This series is my attempt to capture the advice I wish I'd had early on—backed by authentic stories, not theory.

I've been Head of Product at an acquired ad tech startup (Beachfront), Principal Product Manager at a leading video SSP (SpotX, also acquired), and now VP of Product at NBCUniversal, responsible for all programmatic advertising products powering Peacock and NBCU properties. I also built and sold the ad tech newsletter you're reading now.

If you're early in your ad tech career or trying to level up, this is for you.

OK, that's enough tooting my own horn. Let's dive into today's story.

Ad

When you find what you want, create your own opportunity to get there.

When I began my career in digital advertising, I had little more than an entry-level web design college course to my name and a burning desire to do anything other than attend the law school I was set to start that fall.

So I took an inside sales job at the beginning of summer with a salary based primarily on commission at a company that did something called "ad tech." Worst-case scenario, I could start law school classes in the fall.

I knew within a week that I would hate this job and never last.

God bless all you salespeople out there, but the day-to-day grind of basing your livelihood on convincing others to buy what you're selling takes a particular kind of grit and unwavering optimism that I knew I could never manifest.

But during that week I spent on a beanbag with a laptop scalding my crotch in our only American engineer's townhome (we didn't have an office yet), I became fascinated with the technology behind what I attempted to sell. It was little more than a video player with ads and syndicated content at that point, but the fact that a video player, placed on a website, could generate money out of thin air made my mind spin. Our CEO also extolled grand aspirations for something called an "SSP," a video syndication platform, and even a video app builder platform.

I then learned about the companies and platforms behind the technology serving these ads, like Adap.tv, LiveRail, and an entire ecosystem of companies and advertisers using something called "real-time bidding" to try to win the right to show a video ad on a website.

I was like a caveman discovering fire, except this was real-life money being created out of nowhere, as long as you find the eyeballs to see these ads. I decided then and there that my interests lie on the technology side of advertising, not the sales side.

I had to learn more about how all of this worked, but there was one big problem: That wasn't my job.

How was I going to learn how this stuff worked if I was also expected to hit a certain sales quota? If I didn't perform in my sales role, I would lose my job and, consequently, the opportunity to learn more about the inner workings of the technology behind what I was selling.

I didn't need permission. I needed a path.

Phase 1: Identify the opportunity

So I developed a plan and figured out what type of role would be a better fit for me. I needed a role that would allow me to work more closely with engineers and eventually create the things we sell. This is how I came across the concept of a "product manager."

I found out that a product manager works with the business to determine what customers need, then defines that vision into a set of requirements for engineers to develop. The exact nature of product management changes from company to company, but product managers are ultimately responsible for the success of a given product or feature.

Since my interest lay in the line between how software generated money and how the technology worked, this sounded right up my alley. I devoured any resources I could find on product management to make sure this was what I wanted and, eventually, how I could become one.

While still cranking away on the sales side, I noticed that our CEO was constantly pulled in different directions by the demands of growing a startup and our offshore engineering team. He was acting as a de facto product manager and trying to run a company. So I figured he wouldn't mind if I helped in any way I could, as long as I was pulling my weight on the sales side.

If I couldn't change my title yet, I could change my impact.

Ad

Phase 2: Make yourself useful

This also led me to part two of my plan: volunteering to bring rigor and organization through project management, something I learned in my product management studies. I found out that most companies use software like JIRA to organize their product development and Confluence to organize their internal documentation, and they implemented some weird thing called "scrum" to keep engineering teams accountable and to deliver on roadmapped engineering tasks.

I was all in on becoming a product manager in ad tech. I let law school know I was out and took my chance on pursuing a new path that actually excited me. Now I just needed to finalize my plan on how to make it happen.

The fastest way out of a role you don't want is to dominate it.

Phase 3: Become indispensable

I implemented phase three of my plan when I decided to become the very best salesperson possible, so I would never have to do sales again. I shit you not, I think I emailed every single website and publisher on the internet over the course of a year. I used every tactic I could to find emails and phone numbers for decision-makers at publishers I had a realistic shot at landing, and I did this every day for over a year.

I started making money. A lot of money. Enough money that when I finally went to leadership about transitioning into a product role, I'm sure they were relieved. Now they could adjust the commission model for the new salespeople that they told me I had to train and manage, even after I became what I told them I wanted to be: a product manager.

In my off-hours, I continued studying product and project management and applied what I learned at work while scaling and managing a sales team:

• Implementing agile principles

• Rolling out JIRA for project tracking

• Launching Confluence for internal documentation

• Running scrum ceremonies with dev teams

• Writing product requirements and user stories

I also started devouring books on information architecture and UX design. I watched YouTube videos that taught me how to use graphic design software so I could turn the platforms in our CEO’s head into mockups, and eventually into real, working products.

Eventually, once the sales team hummed along, the company recognized that all my time would be better served managing all our products and eventually promoted me to Head of Product.

Master plan accomplished.

Main lesson: Find a way to create your own opportunities in the situation you're given.

It would have been a tough sell to find a product management job with no relevant college or career experience, so I found a way to get the job I wanted and also made some good money along the way. I might not have had the grit to be a salesperson, but I did have the grit to become a product manager.

I also created this opportunity by choosing to stay at a tiny company; a maneuver like this would have been tricky to pull off if we didn't have a CEO filling multiple roles.

Additional lessons

1. You can teach yourself almost anything if you actually care.

I did this before AI existed, with nothing but the good old-fashioned internet. Today, there's no excuse not to learn fast.

2. Be realistic but relentless.

I couldn't quit sales because I needed to eat. So I learned at night and on weekends. Passion doesn't replace rent, but it can coexist with it.

3. Be extremely useful.

I didn't ask for permission; I solved problems. When you create visible value, companies give you room to grow.

This post is part of Surviving and Thriving in Ad Tech, a series on building a durable, fulfilling career in an industry that rarely slows down.

EVENTS

Marketecture Live III

New speakers just added
Ready or Not: AI Is Changing Advertising

How Agencies Are Adapting to the AI Meteorite 
Bob Lord, Horizon Media & Obele “Brown-West” Hinsley, Colle McVoy

Consumer Protection in the AI Era 
Mark Meador, Federal Trade Commission & Alan Chapell, The Monopoly Report

Sponsored by Experian Marketing Services

Programmatic media buying isn’t getting any easier. Buyers face rising fragmentation of signals and identifiers, shrinking margins, and growing pressure to perform in a privacy-first world. Add in fragmented tech, shifting privacy regulations, and inconsistent identity signals, and reaching the right audience at scale has never been more complex (or costly).

Meanwhile, advertisers demand more transparency, publishers want control, and consumers expect relevance without feeling watched. Open exchanges, legacy data management platforms (DMPs), and spray-and-pray strategies just aren’t cutting it.

That’s why curation has become a trending topic in the ad tech space. In programmatic advertising, curation refers to the strategic integration of enriched data and quality inventory into a single package, often executed as a private marketplace (PMP) deal. This approach reshapes how campaigns are built, activated, and optimized by enabling performance-ready buys that reduce waste, increase relevance, and give marketers greater control over outcomes.

It’s no wonder then that curation investments are accelerating, PMPs are gaining traction, and buyers are shifting budgets accordingly. Over 66% of the $150 billion+ open exchange programmatic market now flows through curated PMPs, with industry giants like Google publicly backing curation as a core strategy.

The Ad Tech Explained team developed this explainer, in partnership with Experian Marketing Services, to explore programmatic curation and why it is becoming essential for privacy-conscious, performance-driven marketers.

Sponsored by Experian

What does programmatic curation mean in ad tech?

Programmatic curation is the process of filtering, structuring, and packaging enriched data with high-quality programmatic inventory, enhanced by real-time optimization to create more targeted, efficient, and transparent media buys. Instead of buying at scale and hoping for the best, curation prioritizes accuracy and value over volume, resulting in more targeted, efficient, and transparent media buys.

As privacy regulations reshape targeting, curation is becoming a key differentiator in an increasingly crowded ad tech space. It connects robust data with quality supply, helping marketers stay effective while giving publishers new ways to monetize with control.

But not all curation is created equal. For best results, programmatic curation requires three core elements:

1. Unique data: Unique, privacy-compliant, and valuable.

2. Strong supply connections: Access to quality inventory from publishers at scale

3. Optimization tools: To measure, refine, and improve performance during a campaign and improve performance throughout the campaign lifecycle.

Why does curation still matter with cookies sticking around?

While third-party cookies and mobile IDs like IDFA haven’t gone away, the industry has already pivoted in anticipation of their decline.

Even without Google flipping the switch, signal fragmentation is here thanks to Apple’s ATT, tighter compliance rules, and new tracking restrictions. Marketers are dealing with less data, more fragmentation, and fewer reliable IDs. So if you want to run an effective, scalable media campaign today, you need tools that go beyond third-party cookies.

That’s why programmatic curation is gaining ground as a better, more sustainable way to buy media. Investments in programmatic curation are growing, and not just for identity resolution. Marketers are leaning in because it:

• Combines verified audience data with premium inventory

• Supports ID-free or context-aware targeting

• Enables scalable, identity-agnostic strategies

• Improves data quality through enrichment

• Creates more controlled, pre-approved media paths via PMPs

Experts anticipate that, at some point, publishers will have to lean into this method. Andrew Casale, CEO of Index Exchange, said, “Curation will be bigger than header bidding and as big as programmatic or RTB — that’s our bet.”

What are the benefits of curation — and who wins?

Curation is one of the few innovations in ad tech delivering meaningful wins across the ecosystem. It helps brands, agencies, and media buyers drive stronger performance, gives publishers more control, and creates a more respectful experience for consumers. Here’s how each group benefits and why it matters.

Brands and agencies: Curation helps brands and agencies run more efficient campaigns by connecting high-quality data with premium supply. It delivers:

• More efficient supply paths with less waste and better performance

• Lower costs than DMP segments and open exchange buys

• Cleaner, pre-vetted inventory that aligns with audience and brand goals

• Future-proofed buying via cookieless and log-level data integrations

• Stronger targeting and measurement driven by enriched data and actual usage signals instead of modeled assumptions

• Better outcomes through real-time supply and data optimization

Publishers: For publishers, curation makes inventory more addressable without giving up control. It enables:

• Better packaging that aligns with buyer needs and campaign goals

• Higher CPMs by curating inventory above open exchange floors

• More control over how audiences are accessed and how inventory is monetized

• Access to higher-value demand through curated, data-backed deals

• Better protection of proprietary audience in a privacy-conscious environment

Consumers: At the end of the chain, curation improves the ad experience for those who matter most: your audience. It supports:

• Higher-quality content alignment for more natural, less disruptive ad experiences tailored to their interests

• Less invasive tracking, as targeting becomes more data-smart and privacy-aware, with reduced reliance on legacy identifiers and personally identifiable information (PII).

Sponsored by Experian

What’s Experian’s role in powering curation?

With the acquisition of Audigent, Experian is now more than just a premier data provider. It’s also a full-service curation partner. Together, the companies deliver end-to-end programmatic curation across data, inventory, and optimization, helping brands and publishers achieve more scalable media strategies.

Experian and Audigent make it easier to activate high-performing, privacy-compliant campaigns at scale with:

• End-to-end curation across data, inventory, and real-time optimization

• Pre-packaged and custom PMP deals with built-in performance signals

• Scalable, privacy-first media solutions aligned to brand objectives, verticals, and KPIs

Curated marketplace deals are already delivering measurable gains. In OpenX, Audigent-curated campaigns increased bid competition by 20% and drove a 118% spike in impressions won over a two-month period.

What does successful curation look like in action? What’s really possible?

Here’s a example for you: A leading homeopathic brand, Boiron, partnered with Audigent to scale customer acquisition for its flu-relief product without increasing media investment. The brand wanted to reach new, privacy-compliant audiences while improving CPA across display and video. Using Audigent’s curated SmartPMPs and CognitivePMPs, the campaign delivered measurable performance improvements:

• 80% lower CPA on display than historical benchmarks

• 40% lower CPA on video

• 30% reduction in data costs

• Significant media efficiency gains through real-time supply and data optimization

• Fully future-proof targeting with zero reliance on cookies or MAIDs

Boiron’s success story is just one example of how curated deals backed by Experian identity and Audigent’s optimization technology can deliver efficiency and impact in a privacy-conscious environment.

About Experian and Audigent

Together with Audigent, Experian’s curated solutions go beyond basic packaging. Every deal is designed for performance, privacy, and relevance, so marketers can activate their media with greater confidence:

• Deal ID-ready audiences for fast, turnkey activation

• Performance-specific curation, including viewability, CTR, and outcome-based targeting

• Custom audience + inventory packages tailored to campaign goals

• Vertical-specific curation for industries like auto, retail, CPG, and financial services

• Real-time optimization signals embedded into every curated deal

With Experian and Audigent, marketers are not just getting better data but also a partner that helps activate it with accuracy, privacy, and performance at scale.

Sponsored by Experian

Sponsored by Mula

Publishers are entering one of the most consequential periods in the history of the open web. Traffic is declining as audiences shift toward walled gardens and AI-driven search experiences that summarize content rather than send users to it.

Operating costs continue to rise, and the long-standing assumption that “more content equals more revenue” has stopped holding true. As AI accelerates, so do the fears: shrinking visibility, unpredictable referral patterns, and a growing sense that the tools available to publishers lag far behind those used by platforms extracting the most attention.

But there’s another way to look at the moment. The same technology that is destabilizing the open web can also be the catalyst that helps publishers reclaim value. If embraced deliberately, AI can unlock more revenue per session, strengthen editorial impact, and modernize the broken growth model that has held the industry back.

The Ad Tech Explained team developed this explainer, in partnership with Mula, to shed light on the ways in which publishers can be using AI to fight AI.

Sponsored by Mula

Let’s start with the basics. What exactly do we mean by “AI-powered publishing”?

AI-powered publishing refers to a shift from manual, reactive optimization to an operating-system-level approach that uses machine intelligence to elevate every page. Historically, only walled gardens had access to systems capable of predicting behavior, improving UX in real time, and optimizing for revenue across millions of surfaces.

In an AI-powered publishing model, the technology acts as a co-pilot: analyzing content, guiding decisions, improving discovery, and ensuring monetization strategies align with actual reader intent. Crucially, this approach doesn’t replace editorial judgment. It amplifies it, bringing goal-based intelligence to workflows that have long relied on guesswork and limited resources.

What does AI-powered publishing look like on the content side?

On the content side, AI evaluates every article through a semantic lens to understand intent, relevance, and the behavioral patterns likely to follow. Systems can identify where readers might want deeper context, where commerce opportunities naturally fit, and where additional content should be injected to create a smoother, more engaging journey.

Modern publisher OS platforms, like Mula, use modular AI agents trained to ingest content, analyze metadata, understand tone and relevance, and generate editorial-safe recommendations. This enables publishers to extend the life of articles, surface the right products and links, and deliver richer, more personalized experiences without adding work for already overstretched editorial teams.

Content is the fun part. But ultimately, publishers need to stay in business. So, what does AI-powered publishing look like on the monetization side?

The monetization layer is where AI can drive some of the most immediate and measurable gains. AI-powered systems dynamically evaluate whether a page should surface affiliate products, programmatic placements, next-page modules, commerce blocks, or incremental scroll units, and in what order.

Instead of setting static rules, agents can continuously test variations, optimize placements, and target toward specific business goals like higher RPS, more affiliate clicks, or deeper scroll depth. Measurement agents track performance in real time, right down to widget views, product clicks, scroll patterns, and revenue contribution. This helps publishers to turn every article into a living asset that improves over time.

Speaking of agents, how does AI-powered publishing intersect with AdCP?

The Ad Context Protocol (AdCP) provides a structured way for publishers to communicate detailed context, goals, and constraints to buyers and AI systems. As agentic advertising grows, these signals become essential: AI systems need richer metadata to understand what content represents, how users are engaging with it, and what outcomes matter for that particular impression.

AI-powered publishing expands the value of AdCP by generating high-quality contextual data in real time and packaging it in a way that programmatic buyers (and future buy-side agents) can act upon. The Mula team is actively exploring and contributing to AdCP’s development, ensuring publishers can translate this emerging standard into better monetization and a stronger competitive position.

Let’s get more specific. Where does Mula fit into the broader shift toward AI-powered publishing?

Mula is an AI-powered operating system designed specifically for publishers. It deploys in minutes via a single tag, requires zero engineering lift, and adapts to existing editorial workflows. Once live, Mula’s modular agents (named after surfing legends) analyze content, generate contextually aligned UX modules, inject commerce or ad placements, and measure performance continuously.

• Agent Hunter evaluates content and identifies optimal injection points.

• Agent Nique generates editorial-safe recommendations and related products.

• Agent Occy optimizes monetization decisions (whether to show ads, products, or next-page links) based on real-time analytics.

• Agent Taka handles dynamic distribution through scroll extensions and widget deployment.

• Agent Andy tracks performance and pushes natural-language summaries i.e. reporting back to publisher teams.

Collectively, these agents give publishers access to the kind of intelligent, goal-based optimization that walled gardens have benefited from for years.

What real-world results are publishers seeing with Mula?

Publishers using Mula are seeing meaningful, measurable improvements within days. ON3, for example, was experiencing declining CPMs on long-form articles. By deploying Mula’s SmartScroll widget and a custom affiliate feed, ON3 unlocked a double-digit lift in revenue per session and more than $3 eCPM on commerce modules, while also increasing scroll depth. All of this was achieved with no dev time and no added editorial work.

Brit + Co saw similarly compelling results. Mula delivered a 200 percent increase in affiliate clicks without requiring any changes to editorial workflow. These outcomes reflect Mula’s ability to find untapped value inside existing content that would otherwise remain dormant.

What should publishers be doing right now to prepare for an AI-powered future?

Publishers should begin by ensuring that pages are tagged and structured in ways that AI systems can interpret, making it easier to extract meaning and optimize for outcomes. They should also evaluate how their existing affiliate, commerce, and content workflows align with AI-powered recommendations, particularly where automation can remove manual bottlenecks.

Also, think like a marketer. Understand more about your users, their value (LTV), and their journeys across your site. Buying traffic will become more important, but you need to know the value of your users’ sessions in real time in order to manage profitability. Be thoughtful with push notifications and partner with companies like Pushly to understand what content should be pushed to which cohort and when. Publishing newsletters (and working with companies like Zeta to monetize them) becomes more important, and this also becomes a vehicle for retention.

Most importantly, publishers should treat AI not as a threat, but as a performance partner. Systems like Mula allow teams to drive more value from the content they’re already producing, modernize monetization without adding operational strain, and remain competitive as walled gardens and AI search reshape the ecosystem. Preparing now positions publishers to benefit from the next evolution of the open web rather than be displaced by it.

About Mula

Mula is headquartered in Mill Valley, CA, and backed by Offline Ventures. It provides an agentic OS that brings scale and efficiency to digital publisher monetization. Mula is a system of AI agents that analyzes content on a particular URL and user behavior to inject contextually relevant products and a publisher’s own content into widgets on that page. Much like today’s social media platforms, the goal is to keep users dwelling, scrolling, and transacting. By embedding into tools publishers already use, and extending this approach to platforms like Microsoft Teams, Mula lowers barriers to adoption so it can be a natural extension of a publisher’s workflow, not another system to learn.

Sponsored by Mula

In recent years, several independent companies have vied to be the go-to performance CTV platform. Companies like MNTN, Vibe, and tvScientific look to bring the same type of performance advertising, focusing on outcomes, that’s long been available on search and social to the living room screen.

While DSPs like The Trade Desk or DV360 cater to the largest agencies and advertisers, niche performance CTV advertising DSPs serve small and medium-sized businesses that are less interested in branding exercises typical of the largest advertisers. Instead, they provide access to the same high-quality CTV inventory, typically at lower minimums, while focusing on outcomes and performance.

While these businesses continue to grow independently, a pure-play performance CTV DSP is an alluring acquisition target for a social media company with an existing self-serve advertising business that wants to extend its reach to CTV — someone like Pinterest.

Pinterest has agreed to acquire the performance CTV advertising platform tvScientific for an undisclosed sum (one we know is north of $125 million, since the deal requires regulatory approval). When it comes to performance CTV advertising platforms, there are only a handful of pure-play assets available — and Pinterest just took one of them off the board.

Pinterest brought in a smidge over a billion dollars in revenue in Q3 2025. While it doesn’t break out advertising revenue specifically, it’s assumed that most of these earnings come from advertising.

However, the company remains well behind its social media brethren, with 600 million monthly active users (compare this to Snapchat's 943 million MAUs and $1.5 billion in revenue or Meta's 4 billion+ MAUs and $51 billion in revenue).

But what it lacks in users and rev, it makes up for in unique offerings. Bucketing Pinterest into “social media” may not be a fair comparison at this point, given how different its offering is from those of companies sometimes mentioned in the same breath. People flock to Pinterest for creative inspiration, home decor ideas, fashion, and more.

A main interaction method with the platform is search, where a user types exactly what they are looking for, including potential products to purchase. This means the company holds valuable search data and serves as a gateway for purchase intent. Pinterest views the marriage of this unique data with tvScientific’s platform as a win:

Pinterest could integrate tvScientific in its existing offering to extend the reach of Pinterest advertisers to CTV. But Pinterest will need to leverage tvScientific’s existing performance advertising capabilities to clearly demonstrate the value of CTV advertising, as existing Pinterest advertisers may not be ready to pay the elevated cost of high-quality CTV inventory.

Pinterest doesn’t own the CTV inventory accessible to tvScientific (compared to the inventory they sell on-platform today), so it would need to demonstrate clear value for any markup applied between inventory acquisition and the bid prices from its advertisers.

If Pinterest wants to leverage its existing audience data on CTV, it needs to elegantly bridge the gap between Pinterest users and CTV viewers across the various inventory sources on which tvScientific serves ads.

Marketecture Live III

Marketecture Live: Consumers in Control Speaker Announcement

Marketecture Live is designed for the doers building what comes next and the thinkers shaping how it all works. We’ve announced our first slate of featured speakers, with more to be announced in the coming weeks.

Early Bird pricing ends January 6, 2026: secure your tickets now before rates go up.

In CTV targeting, an easy yet blunt approach is to bridge that gap using IP addresses, but this path can result in serving ads for the dresses a daughter has been shopping for during Dad’s football game. All devices in the house share the same IP address, which can lead to undesirable coarse targeting issues.

So while using IP addresses to target isn’t as precise, the upside for a platform operator is that it opens up more inventory, since it pulls in more eligible devices in a given household. It also creates more surface area to credit attribution, which may be a key aspect Pinterest is after.

Performance advertising typically requires advertisers to drop conversion pixels or integrate with a conversion API to send conversion data back to the ad platform. (Pinterest already has a conversion API.)

Pinterest already recommends passing IP address on all conversion events (sign-up, add to cart, checkout, etc.), so I would expect Pinterest to seamlessly integrate any CTV exposure data from tvScientific directly into its existing conversion API for CTV-extended campaigns. This could realize a goal of the acquisition — Pinterest may be after more opportunities to tie an IP address from an ad exposure to a conversion event.

If Pinterest is having trouble gaining credit for its advertising offering, the tvScientific acquisition is a perfect way to capture more opportunities to claim credit for a conversion. If they serve a single ad on a TV screen in a household, then they can now claim credit for any conversion registered from that same household IP address on any other device in the house.

The validity of this blunt attribution approach has always been a semi-contentious debate. In our prior example, if Dad saw a CTV ad for a dress his daughter shopped for, but the daughter never saw the ad and still bought the dress, should it still be attributed to the CTV ad the daughter never saw? It may matter to the advertiser, but the platform can still claim the credit and show what a great job it is doing in driving outcomes.

If Pinterest can demonstrate success with this acquisition of TVScientific, will it start a pattern of additional acquisitions in the performance CTV space? There’s a very limited number of pure-play platforms in this area, and perhaps other social media platforms might want to assess whether acquiring a performance CTV platform aligns with any future ambitions to extend self-serve advertising options into CTV.

It will also be interesting to see whether Pinterest continues to let tvScientific operate independently or eventually absorbs the company into its offering. Unless they see tvScientific’s business as a true growth driver to the bottom line, it adds a confusing side hustle for Pinterest shareholders to analyze. Maybe Pinterest wants to maintain an air of stability in the short term to limit employee retention and pacify existing business partner integrations.

CTV supply partners already integrated with tvScientific should view this as a win, as it could open a new revenue stream from Pinterest advertisers, along with increased business with existing and new tvScientific clients looking to leverage Pinterest audience data.

When ChatGPT burst onto the scene, one of the first questions on everybody's minds was what this new paradigm shift could mean for Google's dominance in search advertising. If users turned to their friendly LLM of choice rather than Google search, would that mean the end of Google? 

Well, Google eventually caught up with its LLM, and some may argue that it has surpassed OpenAI with its latest Gemini 3 model. But now there is the thorny question of how to monetize user interactions with Gemini — and the first answer to that question is beginning to show up in the wild:

Twitter tweet

Google first announced in May that it would begin integrating ads into AI mode, and now we are seeing users encounter real-life examples in the wild.

The ads themselves are very underwhelming from an innovation standpoint; they appear just like sponsored search results at the bottom of an AI-generated answer. The added context of a protracted conversation-like experience gives Google even more fodder to target ads more precisely, but bolting sponsored ads onto the conversation rather than integrating them in any new, innovative way is uninspired. But it ultimately might be the right call.

Ad creative innovation might take the back seat to practicality and trust in the near term. Whether the capitalistic machinations of advertisers would surreptitiously influence LLMs has been a key ethical question when considering integrating advertising into LLMs, and the clear separation of answers and ads should assuage any concerns in the near term. 

Google has made its move into advertising within LLMs, and now all eyes are on when and how OpenAI will integrate advertising into ChatGPT. It sounds increasingly unlikely that OpenAI will weave advertising into ChatGPT's conversations with users. Altman looks to be angling for a more elegant commerce monetization opportunity:

He clearly thinks advertising corrupts the experience, but he stopped short of saying ads are off the table:

The only other hint of how advertising might manifest within ChatGPT comes from the inaugural episode of OpenAI's in-house podcast:

He then brings the conversation back toward the issue of trust and how advertising corrupts that experience. So what do we know to make a prediction? It’s sounding like the CEO of OpenAI is morally opposed to how advertising can corrupt a product's incentives, but also lauds advertising when it's "cool" or relevant.

One thing most people can agree on is that Instagram delivers extremely relevant ads relative to most other advertising experiences. Like Mr. Altman and many others, I have fallen victim to the extreme relevance of Instagram advertising, resulting in direct conversions for many an advertiser targeting my reels feed.

But my reels feed is reserved for mindless escapism while I use ChatGPT for serious research. While I can fly past an irrelevant reels ad with a flick of a thumb without a second thought, any whiff of ChatGPT steering me toward a product that would financially benefit from my attention would most certainly cause severe erosion of trust. 

Which brings us back to Google slapping a "sponsored" section onto its LLM outputs. While it lacks elegance, it creates a clear separation between organic answers and advertising. Users might not like their experience cluttered with ads, but at least they know their conversations with Gemini are uncorrupted and free of any monetary incentives that would sway them in a particular direction. All of that lives in the "sponsored" section. 

Marketecture Live III

Marketecture Live is going BIG with our new partners: Adweek and TVREV.

Marketecture Live III: Consumers in Control takes place on March 10-11, 2026, at the Glasshouse in NYC.

• 2 Full Days

• 3 Tracks

• 1,000+ Attendees

• Celebrity Speakers

Public markets might dictate how aggressively each company injects advertising into its chat experiences. 

Even though OpenAI remains private, many predict an IPO in late 2026 or early 2027, with the company targeting a $1 trillion valuation. There are 800 million weekly ChatGPT users, but only an estimated 5% pay for the service. OpenAI will, at the very least, have to put out a plan to monetize the 95% of its users who are free to justify a $1 trillion valuation.

The AI market froth has pulled back recently, and Google may feel pressure to demonstrate to the world that it's ready to shift its search revenue toward evolving user preferences for conversational LLM experiences. The company’s stock has shown remarkable resilience despite a market growing more fearful of an AI bubble, and Gemini's widespread rollout of advertising can assure investors it's ready to capitalize on a transition from search engines to answer engines.

OpenAI does have the luxury of seeing how Google's "sponsored" ads strategy plays out over the next year.

Even if it turns out to be the correct path, OpenAI can’t just flip a few switches to redirect a firehose of millions of individual advertisers from its search advertising business to its new AI product. OpenAI will have to build an advanced self-serve advertising platform on par with Google's offerings if it hopes ever to compete — a monumental task that would take years to realize fully. 

Sam Altman claims that OpenAI will hit $20 billion in revenue this year and is on track to reach hundreds of billions by 2030. Compare that to Google, which hit $350 billion in revenue in 2024. It's unclear whether Altman is factoring in advertising revenue potential into his future revenue projections, but he will have to figure out how to ramp revenue to pay for $1.4 trillion in infrastructure commitments. 

On the Bg2 podcast, Brad Gerstner recently asked Altman how he plans to pay for infrastructure commitments given his current revenue, and this didn’t make Sam happy. His defensive answer elicits even more questions, but it most likely demonstrates a CEO who knows more than the people asking the questions (and a man morally conflicted with advertising being the most obvious business model).

Subscriptions, API usage, licensing deals, and maybe even hardware will contribute to OpenAI's bottom line, but advertising is the most straightforward path to the hockey-stick revenue growth required in the short term and to help pay its ballooning infrastructure commitments. 

Sponsored by Playwire

Digital advertising has come a long way toward cleaning up its act. The industry has matured, regulation has evolved, and many players have worked hard to make transparency and quality measurable. But there’s still work to do.

Today, publishers operate under an extraordinary level of scrutiny. Every impression is evaluated. Every metric is dissected. Made-for-advertising (MFA) sites have cast a shadow over legitimate publishers, forcing even the best to constantly prove their worth. That’s not necessarily bad. Quality matters, and bad actors should be exposed.

But while publishers are held to the highest standards of transparency, much of the ecosystem continues to operate behind a curtain. The Ad Tech Explained team developed this explainer, in partnership with Playwire CEO Jayson Dubin, to shed light on the transparency gap that still exists across the digital supply chain, and what needs to happen to close it.

Let’s start at the beginning: The digital advertising industry has made great strides when it comes to improving transparency, but there’s still a ways to go. Where have we seen the most progress, and where does the industry need to focus more attention?

In the past several years, heightened scrutiny of publisher practices has significantly cleaned up the market. Frameworks around quality, performance, and transparency (QPT), along with accountability standards from organizations such as Jounce Media, have pushed publishers to modernize their operations.

These efforts have worked. Publishers who embrace transparency have seen improved monetization, stronger relationships with advertisers, and greater trust across the ecosystem. But this intense oversight has not been matched by reciprocal accountability on the buy side.

That imbalance is where the transparency gap begins.

Let’s get a little more specific. What’s missing?

Critical transparency gaps remain, including hidden fee structures, lack of visibility into what each player in the ecosystem paid and is subsequently charging for an impression, and anonymous buyer behavior that enables malicious ads. Consider:

Purchase price versus final bid. Many SSPs pass along a bid and expect publishers to trust that it’s fair. But publishers often have little visibility into what the SSP originally paid for the impression, making it nearly impossible to verify whether the SSP is adhering to the contracted fee structure.

Some partners share this data openly. Others ask publishers to “trust the numbers” without providing verification. That’s not transparency. That’s asking the industry to operate blind.

Universal buyer identification. Publishers spend countless hours tracing the source of malicious redirects or bad creative, often hitting a dead end because the actual buyer can’t be identified. Even after investing heavily in ad-quality tools, many are still playing whack-a-mole because the supply chain lacks consistent buyer visibility.

Transaction IDs were designed to give the buy side a clearer view of the publisher landscape. But there’s no equivalent mechanism for publishers to identify buyers, leaving half the ecosystem in the dark.

Is this opacity happening across the board? Is the entire ecosystem at fault?

According to Playwire, the answer is nuanced. Not every partner is part of the problem, but enough are to make it a systemic issue.

Playwire’s experience with a recent SSP partner highlights the problem. The SSP’s terms were opaque, reporting was limited, and verification was impossible. When Playwire pushed for more insight, inconsistencies came to light and were eventually corrected.

But that discovery raises a larger question: If transparency can suddenly “appear” when scrutiny increases, what was happening before? This pattern suggests that opacity isn’t always an oversight. Sometimes, it’s a choice.

Not every company is part of the problem. Playwire notes that some are setting the bar for what true partnership looks like. Industry leaders such as Magnite, PubMatic, and Index Exchange demonstrate that transparency and profitability are not mutually exclusive. These partners provide clear visibility into fee structures, share purchase price data, and maintain open communication with publishers.

They’re proof that transparency benefits everyone.

When SSPs disclose what they paid for an impression, publishers can optimize yield with precision. When universal buyer IDs exist, bad ads can be traced and eliminated faster, protecting both publisher integrity and advertiser investment. When the supply chain operates openly, inefficiency declines, and total revenue grows.

OK. So, how do we fix it? What needs to happen?

The solution isn’t complicated. It just requires the same accountability on the buy side that publishers already live with. For example:

• Purchase price and subsequent bid reporting should be mandatory. If publishers must prove their value at every turn, SSPs should disclose their take rates and margins. Transparent pricing is the foundation of trust.

• Universal buyer IDs need to become standard. Every buyer should have a persistent identifier across the supply chain. When a malicious ad appears, publishers should be able to trace it to its source immediately rather than spending days investigating.

• Industry bodies need to enforce buy-side standards. The IAB, TAG, and other organizations have spent years establishing frameworks for publisher quality. It’s time those same principles applied to SSP and DSP transparency.

These measures would move the industry from selective transparency to shared responsibility.

What’s Playwire doing to bring this future into focus?

Playwire continues to advocate for transparency that flows in both directions. The company has built its reputation by providing publishers and demand partners with detailed analytics and open reporting, grounded in the belief that knowledge is power. Its recognition as a Bellwether partner by Jounce Media reinforces this approach, validating that operational transparency leads to higher performance and healthier relationships.

Playwire works exclusively with demand partners that maintain clear fee disclosures and provide data access to validate performance. The company also encourages wider industry adoption of reciprocal transparency frameworks.

Ultimately, Playwire’s view is simple: Transparency should not be a competitive advantage. It should be the cost of entry for doing business in digital advertising.

About Playwire

Playwire is dedicated to helping publishers, portfolio managers, and app developers accelerate their business by amplifying ad revenue and operating more efficiently. This purpose drives every decision we make; it’s at the heart of every new tool we roll out.

For more information, visit www.playwire.com.

Earlier this month, I attended the Agentic Advertising Standards & Community launch presentation. I sat back and absorbed the pitch for what could ultimately develop into a new form of ad buying. The presentation is now available on YouTube for all to see and either laud or criticize.

The presentation was compelling enough that I, like many others in the industry, will have to decide whether agentic media buying is worth dedicating time, brainpower, and resources to in the short term.

With so much constantly changing in digital advertising and AI advancements, knowing where to devote your time is more crucial now than ever. Is this the next revolution in digital advertising or a waste of time like the Google Privacy Sandbox? In any case, it is prudent to approach anything that claims to upend digital media buying with some skepticism while remaining open-minded.

What is AdCP?

The presentation made the case for agentic buying via a new Ad Context Protocol. From the new protocol's GitHub:

Ad Context Protocol (AdCP) is an open standard for advertising automation that enables AI assistants to interact with advertising platforms through natural language.

A collection of companies is forming a non-profit entity to govern AdCP and the standards spun out of the initiative. Curiously, the IAB Tech Lab is not involved at all. On the group's website, they clearly outline the goals of AdCP:

Agentic Media Buying

AdCP looks to facilitate agentic advertising or agentic media buying. If you think about media buying today, there are really two paths: direct and programmatic. 

Direct buys are either set up by advertisers on self-serve platforms or by a publisher on their own ad server on behalf of an advertiser. 

Programmatic buying automates this process by using specialized systems (SSPs + DSPs) to connect inventory to demand and facilitate real-time auctions. 

While technically a form of direct buying, you can consider buying anything on the walled gardens as another form of media buying that requires specialized skill sets and API integrations. 

Each form of buying has its pros and cons. 

• Direct buying ensures no unnecessary ad tech taxes, but it requires human communication and can be hard to scale across publishers.

• Programmatic offers scale and automation, but it can lead to unnecessary fees from intermediaries.

• Walled gardens all have their own way of doing things and can require bespoke integrations, which can be burdensome. 

Agentic buying could extract the best parts of each of these buying paths. It could allow advertisers to automate media purchases via natural language powered by large language models. Here is an example prompt from the AdCP website that a buyer could plug into a buyer agent:

A buyer agent could interpret the buyer's natural-language input and interact with various seller agents to accomplish that goal. But what about activation? How do these campaigns actually get set up and run?

Which brings us to perhaps the most interesting portion of the AdCP launch presentation: a demonstration (22:09). The demo shows a buyer agent, powered by Scope3, purchasing inventory from a publisher (LG) via a seller agent, powered by Swivel, and activating a campaign in the SpringServe ad server. 

It's one thing to talk theory, but this demo clearly illustrates what agentic buying is or could be. In the demo, LG has "products" set up in Swivel that let them define a set of inventory available for purchase, including characteristics such as CPM, format, guaranteed vs. non-guaranteed, etc. Think of an ad "product" like a multi-buyer PMP deal in programmatic.

The Scope3 buying agent uses AdCP to interact with the Swivel seller agent to discover relevant ad products based on the buyer's input. The buyer can then ask the Swivel seller agent to execute a buy using one of the listed ad products and to push a proposal to the seller, who will appear in the Swivel UI for a human to review and approve.

Now we arrive at activation. Once the proposal is approved, Swivel then uses its preexisting API integration with SpringServe to create the necessary objects in the ad server to execute the buy (campaign, line item, targeting, etc.).

Do we need agentic buying?

This is the central question we should all be asking. The presentation highlighted what sucks about each current form of media buying. Let's take a look, and I'll provide some thoughts.

Direct

Requires sales teams to sell and humans to traffic campaigns, which limits scale.

True! It is honestly absurd to me that direct buying still exists in 2025. Humans talking to other humans, then pointing and clicking buttons, feels like an anachronism that we should relegate to history.

Walled Gardens / Ads APIs

Bespoke APIs require custom integration, which requires resources and limits scale.

This is pretty much only applicable to the walled gardens like Meta or Google. Maintaining any custom integration is annoying enough; now multiply that by the number of ad APIs you must integrate with. 

My question is: Are we really banking on these walled gardens integrating with AdCP? If they don't build out seller agents to facilitate agentic buying, then we are stuck with the bespoke integrations. 

Programmatic 

1. Difficult to leverage publisher first-party data on buys

2. Less spend goes toward working media 

3. Cannot access the custom capabilities of publishers

The presentation spent more time bashing programmatic than any other form of buying. Coincidence or intentional? Comments on each point:

1. Publishers can layer first-party data on both PMPs and PGs; however, this requires human coordination. 

2. The biggest knock on programmatic. Unnecessary intermediaries in the mix or high fees take away budget that buyers could dedicate to working media. But Swivel, Scope3, and anyone powering an agent will also need to be paid. Are we swapping SSP and DSP fees for Seller and Buyer agent fees?

3. If publishers offer custom units with no defined programmatic standards, then yes, this is an issue. But the most popular ad products gain traction in IAB Tech Lab working groups that develop standards. A current example is the ongoing development of programmatic pause ad standards. But I will admit this process is long and sometimes painful. 

Agentic buying could solve some of these problems, but questions remain. The demo is slick and all, but when put into practice, will agentic buying offer savings in time and resources to warrant adoption? Will more spend ultimately go toward working media, or will it merely be shifted to alternative players?

My initial feeling is that some form of agentic buying should ultimately supplant direct advertising or programmatic guaranteed. If AI agents can automate negotiations and painful setup processes for simple buys that may use publisher first-party data, that is a no-brainer.

But can agentic buying ever really replace programmatic advertising (at least any time soon)? Sure, it has its issues, but programmatic offers a system that is proven to scale, gives advertisers control by letting their buys ultimately live on platforms they control (DSPs), helps them find the lowest price for inventory, and is a proven ecosystem of features, companies, and products.

For now, I view agentic media buying as an experimental alternative to traditional direct buying paths. Something that could eat some direct or minimal programmatic budgets for testing purposes. If it's easy to set up an MVP for both buyers and sellers, why not open an additional demand channel? But the level of difficulty to test is still an open question.

The bull case would be replacing direct in the short term and possibly eating into programmatic in the long term. The bear case is that this new form of buying proves ultimately too cumbersome to use and implement, and does not add enough value to warrant further development. 

Regardless, I'm here for it and will be following along every step of the way.

Last week, Magnite announced its acquisition of Streamr.ai, a company that describes itself as the "easiest way to generate video ads and launch them on CTV in less than 2 minutes." 

According to the press release, Magnite hopes that Streamr.ai technology will help ramp up small and medium-sized business spending: 

Magnite is not the only player in this game, as other demand platforms are salivating at the prospect of onboarding more SMBs to CTV. And is that any surprise, given that Google and Meta have earned billions creating tools that make it easy for SMBs to advertise on their platforms?

But CTV is an entirely separate ball game, and one of the biggest blockers that could prevent an SMB from advertising on CTV is the cost and complexity associated with developing video ads. While developing display creatives requires little more than a nice camera and a basic knowledge of graphic design, creating a video creative worthy of a 4K screen demands more investment and skill.

Ever since OpenAI released its text-to-video AI model, Sora, in February 2024, with other models quickly following, some companies saw an opportunity to use text-to-video AI technology to whip up creatives on behalf of the advertisers that otherwise didn't have the time or money to develop video creatives. 

And now, we have real-world examples of companies cranking out video AI creative. First, it was the big companies like Coca-Cola or Toys R Us dipping their toes in.

And Kalshi reportedly spent just $2,000 to create this unhinged ad that ran during the NBA Finals:

And there are more examples of individual creators producing concepts for brands like Nike or Adidas:

The only thing holding AI advertising back is backlash from consumers. Take one look at the YouTube comment section in the previously mentioned Coca-Cola or Toys R Us ads, and you'll find a sea of backlash against the use of AI for advertising:

Coca-Cola:

• Coming from a billion-dollar company like Coca-Cola, this is an AI embarrassment, and likely a deliberate marketing strategy.

• Nothing like celebrating the spirit of Christmas with the most soulless commercial possible.

• No animators were employed in the making of this advertisement.

• Words cannot describe the level of hatred I have for this ad

Toys R Us

• This is like if I had a nightmare about Toys R Us.

• This commercial killed the magic

• Soulless venture capitalists further ruin Toys R Us by producing soulless ad

• Gosh this is horrendous. I felt nothing watching it (other than a bit sick)

Given that creatives were the first career paths threatened by the generative AI boom, it is no wonder that the first examples of AI-generated video advertisements would receive the most criticism. The harsh reality is that we will most likely view comments like these in the same way as the destruction of textile machinery by the Luddites in the 19th century. 

Generative AI creatives will become indistinguishable from reality, and the cost to produce them will only decrease, which will mean consumers won't even know what to criticize as they eventually flood CTV. 

The prior examples were generated by creative professionals, but with companies like Streamr.ai and others developing more automated solutions, it’s only a matter of time until AI creatives from advertisers of all sizes start hitting the living room.

Big players like Amazon have released new tools to generate video creatives, and independent players like Creatify can also produce video ads based on Amazon product listings. And it's not just full-on AI-generated creative. Spaceback has created tools for advertisers that can crank out CTV ads using existing social assets.

But what does this coming onslaught of generated creatives mean from an ad tech perspective? The one thing I can be certain of is that it will mean more unique creatives flowing through the video advertising ecosystem.

I anticipate car commercials that dynamically swap environments based on a user's location or even advertisements that dynamically insert the products you've added to your cart. Scenarios like this could lead to an onslaught of creatives that publishers must deal with. 

Connect with top brands and agencies at Marketecture Live

Last Monday night, I got hit with this Gillette ad during the Buccaneers vs. Texans game. It is customized to the matchup, with Troy Aikman referencing the Buccaneers, along with the Buccaneers logo and branding in the ad itself. Now, I'm not saying this is an AI-generated Troy Aikman, but it got me thinking about the potential of using content metadata to target uniquely generated creative. 

Perhaps Mr. Aikman filmed 32 separate segments for each NFL team (although I've only found one other for the Cowboys vs. Eagles game, excluding Troy Aikman), but this is a manual and costly process. I anticipate advertisers will start creating variations like this automatically through generative AI or other tools. 

If generated creative variations become more commonplace, there will be an even greater desire for accurate programmatic content metadata and user identity. Advertisers and ad tech providers have always craved these data points for targeting and reporting purposes, but now they can use these inputs to generate relevant creative on the fly. For example, if a publisher passes which game an ad opportunity is for in a bid request, then Gillette can slot in the appropriate NFL team-based ad more seamlessly. 

There could be a world where advertisers customize CTV ads down to the user level, allowing them to showcase products the user has shown interest in or tailor an ad to target their interests and desires more effectively. But the number of unique creatives generated from a practice like this could put an immense strain on publishers who must transcode every variation of an ad. However, the possibility exists and may become an engineering challenge to solve in the future. 

Publishers and ad platforms already have to deal with potentially thousands of creatives a day. What happens when this increases to tens and hundreds of thousands? 

CTV publishers can already enforce creative standards by analyzing the creative metadata and files. Pubs can quickly interrogate things like width, height, bitrate, advertiser, category, and more, so dealing with a deluge of new creatives is more of a matter of scaling existing systems. 

However, what happens when anybody with a credit card and a dream can access premium streaming inventory? Interrogating the content of the advertisement itself may present unique challenges that publishers will eventually have to deal with. Objectionable, uncanny, or low-quality advertisements may present a jarring experience to users. 

Computer vision technology can detect inappropriate frames of a video (nudity, hate symbols, curse words, etc.), but determining quality may ultimately come down to human taste. If a publisher opens up access to small businesses without guardrails, they may be in for a rude awakening of the AI slop that starts rolling through their programmatic pipes. 

Publishers may want to deal with AI creatives differently than "professionally produced" advertisements, much in the same way advertisers make a distinction between professional content and user-generated content. We may see publishers requiring AI creative to flow through separate deals that require further review for auto-generated creative to make sure they adhere to creative guidelines.

Additionally, a high amount of unique creatives introduces challenges for server-side ad insertion use cases. In some SSAI setups, the first time a DSP bids with a new creative, that opportunity is "burned" because the creative must be registered in a creative approval queue and transcoded before it is ready to serve. This quirk may be fine in VOD use cases, but in live sports environments, for example, a DSP could bid with the same creative thousands of times all at once, and if that creative is approved, it could be thousands of burned opportunities before the creative is ready to serve. 

SSPs sometimes offer creative pre-ingestion tools to allow DSPs to pre-register creatives before they start bidding with them to avoid these situations. In a world of generative AI, creative pre-ingestion becomes much more crucial. 

If generative AI is the key to making SMB spending take off in CTV, one last thing I'm interested in is the economic impact on the video advertising ecosystem. An increase in demand should push prices higher, but SMBs don't have the same deep pockets as large advertisers or agencies. The reality is that these smaller advertisers are likely looking to come in at a much lower price point, which relegates them to remnant or low-quality inventory. 

I anticipate that we will first start seeing SMB AI ads on FAST channels where supply is plentiful. That's not to say that the big streamers won't have their fair share of ads created using generative AI, but it will be in the form of big advertisers streamlining their production process through AI, most likely with much higher quality results. 

Agentic AI could fundamentally change how you interact with ad platforms and perform your job. Whether you work in ad ops, account management, sales, or strategy, if your job requires you to log in to an ad platform to perform a task, then you should understand what agentic AI means for you.

We adopt technology when it enables something new or helps us accomplish something more efficiently. This has been the case from the printing press to the iPhone. After speaking with product leaders and builders across the industry, I can tell you that agentic AI will enable new use cases and increase productivity in ad tech.

Agentic AI, coupled with generative AI, could boost human productivity to levels previously unimaginable. This technology can multiply the output of a single worker many times, which is a daunting prospect for the future job market but also a wondrous opportunity for those who learn to harness its power.

I wrote this article to learn more about agentic AI, understand what's behind the buzzword, and share with you what I learned. I interviewed founders and product leaders who are riding the agentic AI wave. These are the people who must figure out how agentic workflows enhance their existing platforms or how to build entirely new companies based on agentic products.

But before we go any further, we have to learn what agentic AI means.

What is agentic AI?

Agentic AI is the ability of a system to take independent actions toward goals, operating proactively rather than just reactively and adapting its behavior as needed to achieve them. An AI agent is a tool or system that interacts with users, software, or data to accomplish specific goals autonomously.

Agentic AI is like a person's ability to plan and act independently. An AI agent is the person applying those skills to get work done and achieve specific results.

It also helps to compare this to generative AI. Generative AI is a type of AI that creates new content—like text, images, code, or audio—based on patterns it learned from training data. Generative AI is a capability that agentic AI may use to produce outputs as part of achieving goals. An agent might use generative AI to write a personalized email or summarize a meeting.

The real magic is when all these AI concepts come together. Generative AI text output ushered in the ability to interface with an application via chat, which allows a user to set up agentic use cases via the same natural language input mechanism. It also allows an AI to produce results from those interactions in the form of text, images, or video. This combination of agentic AI plus generative AI is what will usher in a profound shift in the way we interact with technology. 

AI agents can reason and figure out the best way to accomplish tasks for a user, but to do so, they need access to ad platforms and systems and an understanding of their available capabilities. That’s why we now have to touch on MCP. This term came up frequently in all my conversations. 

What is MCP?

Anthropic, the company behind the AI assistant Claude, introduced MCP, or model context protocol, last year in an attempt to standardize how AI assistants connect to other systems. The idea is that a system could host an MCP server that allows AI assistants to understand what tools are available and how they work, along with providing a natural language description of each tool. 

Think of it like API documentation for humans today. APIs allow external systems to push or pull information to or from a platform. But to integrate your system into a platform, you have to understand what endpoints to call, what information to provide, and how to authenticate. This information would all be outlined in human-readable API documentation and implemented by engineers.

A chatbot or AI assistant can query an MCP server and ask for everything it can do, alongside natural language descriptions of what the tool does and the required information to accomplish that task.

Without an MCP or a similar standard, any agentic tool would have to build out a bespoke integration for every app it interacts with using legacy APIs. With MCP, this process could be mostly automated and enable any platform to support agentic workflows, allowing agentic solutions to scale rapidly.

Integrating agentic AI

When I spoke to David Dworin, FreeWheel's Chief Product Officer, he told me he is thinking about agentic AI from two frames of reference:

1. How can FreeWheel better accommodate third-party agents that need to access FreeWheel?

He is thinking about how to enhance FreeWheel APIs or data products to make it easy for agents to interact with the platform. He mentioned that integration with third-party agents could even leverage a FreeWheel MCP server, which his team is considering.

2. How can FreeWheel enhance its system with agentic workflows?

David touted agentic as an evolution of typical workflow automation (if this then that). So he and his team are looking at potential use cases where he could supercharge automated workflows with agentic AI:

He mentioned a potential troubleshooting use case where you can run a diagnostic, and then FreeWheel tools could recommend the action to take or even take that action on your behalf.

I loved this two-part framing, as it establishes a baseline framework to explore agentic solutions in ad tech — not an easy task with novel technology. It also sets us up to explore the different approaches we may see play out regarding model and agent ownership. Let's use David's framing to explore deeper. 

Evolving platforms 

Humans interacting with a user interface has served as the lone method of making an ad platform do something. Agentic AI could shake up this paradigm for the first time in the history of ad tech. Joe Hirsch, CEO of Swivel, who was previously CEO of video ad server SpringServe (acquired by Magnite), put it more directly when we spoke:

Joe is especially invested in this future as his company is looking to power the agents that could serve as members of your ad ops team. Swivel "is on a mission using AI and machine learning to automate media's most complex ad delivery strategies, freeing humans to achieve their loftiest business goals."

This is a lofty goal in and of itself, but Joe thinks Swivel can empower users with the help of agentic AI and unlock possibilities that humans aren't capable of:

AI agents don't sleep; they could have instantaneous access to historical and real-time data, and you can train them on your entire history of changes to ad serving, deal setup, operational policy, and goals. 

An ad ops agent either baked into an ad platform or through a third party like Swivel could make thousands of changes throughout the day based on all these inputs — a truly impossible task for a human. 

But in the meantime, man and machine still must work together — and other platform operators like PubMatic are already launching and testing out new ideas of what this could look like. 

Ankur Srivastava, Vice President of Product Management at PubMatic, told me that the company has released PubMatic Assistant, an LLM-powered tool that is currently in closed beta.

PubMatic has integrated its assistant into reporting tools to extract complex insights quickly, and is now focusing on integrating agentic AI into the deal management lifecycle (demand insights, deal creation, troubleshooting, and optimization). PubMatic Assistant, for example, can now help with programmatic deal troubleshooting.

Ankur observed that troubleshooting skills among ad operations teams can vary. PubMatic Assistant sets a consistent baseline of competency for all customers out of the box.

As someone who has spent 14 years troubleshooting programmatic issues, I've found that there is truly an art to the practice. Whenever I attack a new troubleshooting issue, I have 14 years of experience and pattern recognition to fall back on. But I also don't have millions of programmatic transactions plugged into my brain. When will these tools outperform me in troubleshooting? I fear very soon. 

One advantage I have over AI currently is that humans are still the primary users trafficking campaigns. An AI may be able to tell you that a DSP has a 0% response rate, but if it's not responding, it has no data to work with to troubleshoot the issue. This situation is where human-to-human communication comes into play, which may determine that a human error (like setting incorrect flight dates) is to blame. 

As long as humans configure the operational setup of supply and demand, human workers will remain a vital resource. If agents start taking over both sides, we may see a gradual decline in human necessity and a rapid increase in the need for agent-to-agent communication.

Agentic communication

The topic of agentic communication and interoperability came up in every conversation I had — and for good reasons. 

First, this topic is bound to be top of mind given interoperability among platforms is the heart and soul of programmatic advertising, dating back to the first real-time transaction between platforms — which, curiously enough, I recently learned may have originated from PubMatic. (Thank you for the history lesson, Ari. “Yield“ should be required reading for anybody working in ad tech.) 

Second, agentic communication drastically increases the capabilities of any agentic tool. If agents cannot break free from their systems, then they can only perform actions within a closed ecosystem. 

Outside ad tech, I often hear a recurring example of how I could one day ask my personal AI agent to book me a flight for my upcoming trip. This agent will have access to my calendar to know when my trip is and access to airline booking systems to make the reservation. 

In ad tech, users typically interact with multiple systems, such as DSPs, brand safety vendors, creative ad servers, order management systems, and more. These systems typically already have integrations together, but the usefulness of those integrations could increase significantly through agentic AI.

Let's return to our programmatic troubleshooting example, where a particular programmatic deal booked in a DSP is not responding. An AI agent confined to an SSP may not be able to identify or fix the issue since the problem lies within the DSP. What if our SSP troubleshooting agent could talk to a DSP troubleshooting agent?

Ankur from PubMatic presented me with a potential use case for agentic communication between an SSP agent and a DSP agent.

The SSP agent could inform the DSP agent that the publisher is asking why this deal is not bidding, and the DSP agent could inform both the SSP agent and the advertiser why. Then the humans could take the necessary steps to solve the issue. Further in the future, the agents could have more autonomy to resolve the issue themselves (with guardrails in place).

Integrations among platforms today leverage APIs that allow one system to perform an action on another system. There could be an API action to read or set a campaign budget, deactivate a creative, or update a deal name. Platform operators can define all these actions within an MCP that an external agent can read and understand. It can then interpret a user's input and reason which abilities it needs to accomplish the intent.

Integrating external agents into platforms aligns with another common subject that came up in my interviews: model and agent ownership. While platforms will likely have their own internal AI systems, another school of thought is that publishers, advertisers, or agencies could own or license an internal AI model that will interact with all of their ad systems. 

This model could be pre-trained on all internal brand safety standards, operational processes, and preferences, along with seeding it with baseline ad tech knowledge, including jargon, reporting terms, and understanding of how ad platforms work. The company-level agent would then plug into all the platforms and tools to run an advertising business. It would know exactly what it could do by interacting with the MCP servers of those platforms and tools. 

Joe from Swivel envisions this concept as a possible future component of his company:

He went on to explain how a customer can use natural language to perform ad ops, gather reporting, and do anything from a single entry point.

As an advertiser, imagine you just convinced your client to increase the budget. You could ask your agent:

Find all deals for my advertiser, Pizza Shack, and pause all creatives with a conversion rate by zip code lower than 5%, and shift the budgets allocated for the paused creatives evenly amongst the remaining creatives, and append '— zip code optimized' to the deal name of all updated deals. Then analyze the deactivated creatives and tell me if they offered a promotion and produce a report broken down by state to show me conversions by promotion vs no promotion.

This prompt is an extremely loaded request, but bear with me for the thought experiment. Your agent will have access to the DSP to make creative and deal changes, and internal reporting aggregation tools to pull reporting data. The agent can also access your creative library and use multi-modal visual and text caption scanning to understand if the creative included a coupon code.

Your agent understands exactly which system to interact with to accomplish its goal. But given the complexity of the prompt, it might be a good time to review the concept of "human in the loop."

Humans and guardrails

I don't know about you, but I would not trust an AI at this point to go out and adjust live deals immediately after a prompt. But this is where the concept of "human in the loop" comes into play. After my prompt above, the model could generate a response that outlines what it thinks you want it to do before it performs that action. For example:

Here is what I think you are asking me to do:

1. Find all deals assigned to the key name "Advertiser" = "Pizza_Shack" 

2. Analyze reporting data for "Advertiser" = "Pizza_Shack" broken down by all United States zip codes with "Metric" = "Conversion_Rate"

3. When "Conversion_Rate" < 0.05 for a specific zip code, do not allow that creative to serve that zip code.

4. Create a .csv file with columns = "State", "Promotion or No Promotion" (as a boolean 1 = promotion, 0= no promotion). 

Questions:

1. What date range should I look back to when analyzing the conversion rate for Pizza Shack creatives?

2. You don't want me to pause certain creatives across all zip codes, but you currently don't have any zip code targeting mechanism applied. Would you like me to create a custom rule to exclude the low-performing zip codes from specific zip codes?

3. When analyzing creative video, can you define "promotion"? Does this mean a coupon code is present on screen for any amount of time?

At this point, you would provide further clarifications to your in-house LLM and confirm its approach, with the entire interaction training the model to better understand in the future. After a final confirmation, it would automatically accomplish your intended deal updates, creative scanning, and produce a report for you. Welcome to agentic ad tech.

What if you wanted to take this a step further and enrich your campaigns with first-party or licensed audience data? 

Dstillery, an AI ad targeting company, has already developed for this use case. Dstillery uses artificial intelligence and machine learning to analyze large datasets and create custom audience segments for digital advertising campaigns.

I spoke with their Chief Product Officer, TJ In, and VP of AI Products, Mark Jung, to learn more about how they are incorporating agentic workflows into their products. Dstillery has already stood up an MCP server that defines all the tools available through their platform.

We walked through a demo leveraging these tools where their AI agent (which they referred to as 'DS-1') can query Dstillery's system to build and activate audiences that agencies and advertisers could later use for targeting ad campaigns. In our example, we looked at two separate approaches to identify relevant audiences. First, we looked for people who want to buy bobbleheads, and the tool returned segments of action figure enthusiasts, sports fans, and other groupings of people who the AI thought would buy bobbleheads.

Mark told me this is one of the benefits of AI, where legacy practices would be a human searching a DMP for "bobbleheads" and that DMP only returning segments with bobbleheads in the name or description. But with AI, it can reason the type of people who collect bobbleheads and what kinds of audiences relate to these people. So rather than relying on humans to make these connections and searching for keywords of audiences that could loosely fit your targeting criteria, these agentic workflows can find relevant audiences better than any human could.

In the second example, we tasked our AI agent to look at the first-party data of a large Asian clothing retailer and to return lookalike first-party segments along with relevant third-party segments, and to predict the potential performance lift that these segments would yield if applied to a campaign. 

The tool returned lookalike segments and third-party data segments, including women's fashion brand shoppers, UK boutique women's fashion, luggage shoppers, and non-endemic segments like drama show watchers. The tool then provided the predicted performance lift that each segment would yield for a campaign before an advertiser served a single impression.

TJ pointed out that tasks like identifying the best audience segments, creating custom audiences, and activation could take anywhere from days to weeks, depending on the number of back-and-forth emails and processes involved. When AI agents like DS-1 are available directly in the hands of agencies and media buyers, they can complete these same tasks in minutes with more accuracy.

Job killer or supercharger?

Many discussions about AI and its impact on the economy ultimately drift toward the awkward problem of white-collar job cannibalization — and it's a valid concern in ad tech. If AI agents can traffic campaigns, troubleshoot programmatic deals, produce data analysis, and everything in between, where does that leave us? Will companies need large ad ops teams at publishers, media managers at agencies, or data scientists anywhere?

At this beginning stage, while the AI models are still learning, entry-level jobs are most at risk. Couple that with an uncertain economy due to tariffs, and Gen Z grads are already facing an extremely bleak job market. Employees aged 20-24 occupied 6.5% of all jobs in advertising, public relations, and other related services last year, the lowest since 2020 (down from 10.5% in 2019).

Against the backdrop of this daunting job market, the wealthy person most responsible for the gloomy outlook, Sam Altman, is telling them, "This is probably the most exciting time to be starting one's career, maybe ever."

Despite the tone deaf response, maybe we should listen to the OpenAI chief. Agentic AI opens up some truly limitless possibilities, enabling individuals to accomplish goals that were previously impossible. 

Take the area of data science, for example. Often, data scientists are facing endless queues of requests from different departments. Sometimes companies cannot hire fast enough or keep up with the work required to fulfill the requests.

At various points in my career, I've waited months for a data science request or analysis. But what if anybody had a team of 10 data scientists at their disposal at any point? That's the pitch I heard during my conversation with John Hoctor, CEO and founder of Newton Research.

He told me that he realized that many publishers, agencies, or brands do not have enough data scientists or analysts to keep up with the number of requests they receive every day. This realization led to the genesis of Newton Research, which arms customers with AI agents that supercharge anybody's ability to pull meaning out of media and advertising data. 

His agents come pre-trained on a wide range of analytics, including complex and time-consuming analytics tasks like MMM, incrementality testing, attribution, predictive modeling, campaign insights, and measurement reports. Newton Research enables its customers to "talk to their data" and essentially gives them access to an unlimited number of junior data scientists. So what does this mean in practice? 

Rather than aggregating data, Newton works with the data wherever a customer has it stored, so it doesn't have to leave their preferred system. It allows customers to interact with their data through a large language model and perform agentic actions using that data. 

For example, as an advertiser or agency, you could interface with Newton via chat and ask it to analyze the ROAS of a particular CTV campaign and compare it to past similar campaigns.

To make sure all their customers don't start from scratch, John and the team train Newton on media analytics, advertising metrics, and industry jargon, along with an understanding of integrations — something that could take a junior-level data scientist years to grasp fully. So the model immediately pulls meaning from any request filled with jargon and referencing historical context.

The agentic workflow kicks off once the user provides their goal. The model can reason where it should pull the data from, calculate the ROAS, compare it to past campaigns, and break it down by the highest-performing audience or creative because it knows that's what you like to optimize against. 

From there, the user could choose to leverage those insights into activation. You could ask Newton to create an audience and reallocate budget distribution across creative variations based on performance. The tool could then build a list of user IDs similar to those performing well, redistribute the budget skewed toward the highest-performing creatives, and push all these changes directly to your DSP. 

This example shows how Newton can save media buyers countless hours of work or accomplish tasks they would never be able to achieve alone by leveraging AI agents. Further, it demonstrates how integrating with other ad systems supercharges the capabilities of agentic tools.

There is even development underway between Newton and Dstillery that does precisely this. Mutual customers can use agentic AI to connect Newton's measurement capabilities to Dstillery's audience modeling and activation capabilities to create a closed conversion to activation feedback loop. The idea is that users can find conversion data through Newton, which Distillery could use to generate audience targeting models for activation — a tedious task or outright impossibility for humans on an ongoing basis.

Identifying gaps in the workforce or any area where a human is performing repetitive tasks to free up their time to spend it better elsewhere is something I heard from another founder. 

Adam Epstein is building Gigi, a platform that hooks into the Amazon DSP and serves as an agency's always-on media manager. Users can interact with Gigi via natural language to set up campaigns, optimize performance, or run custom queries for insights. 

Adam gave me a demo of his slick product, and I later heard him on Eric Seufert's Mobile Dev Memo Podcast describe eloquently how he figured out how to build an impactful agentic AI company and product:

Adam identified enterprise media managers as a great place to start, and he pivoted his company to build an agentic AI media manager. Adam went on to explain, "You don't buy software, you hire software."

Companies may soon view AI agents as a supplement to a team rather than a tool that a team uses. Software has the added benefit of never quitting your company and is theoretically infinitely scalable.

Our agentic future

I had always held great envy for those at working age during the dawn of the Internet, a time when you had a vast blue ocean of opportunity literally at your fingertips. At this time, you could leverage an outrageously radical concept, a worldwide interconnected computer network, to dream up infinite possibilities.

I no longer have a valid reason to hold onto that jealousy because we are living in a similar moment. Agentic AI opens the door for founders, product builders, or anybody to imagine use cases that were previously impossible. 

While the threat of human job attrition looms over this groundbreaking set of technologies, those who embrace the coming wave of agentic automation and learn to harness these tools may be living in a time of boundless opportunity that we may never again see in our lifetime.

When I began writing this piece, I had a weak grasp on the concept of agentic AI, and after completing it, I still feel like I have only scratched the surface. But I hope that you, like me, now have a framework to think about agentic AI and how it impacts advertising technology, your company, and yourself.

Now comes the time to figure out collectively how we will integrate agentic AI features and services into our daily workflows. Ad platforms must do the dirty work of figuring out how to properly interoperate with each other — a goal of the Agentic Advertising Collective, organized by none other than Joe Hirsch of Swivel. 

Artificial intelligence is poised to take advertising technology on a ride into a new world of possibilities — but it could also fundamentally change how you succeed in your role.

For every seasoned ad tech professional employing legacy practices, there will be a hungry young worker who will leverage AI to crack into a worrying job market. I recommend you learn how to surf this new wave of AI technology rather than letting the swell sweep you out into a sea of irrelevance.

Sponsored by Sovrn

By 2027, the U.S. commerce media market is projected to top $100 billion, accounting for more than 20% of total advertising budgets. Retail media networks are a major driver of this growth, but they’re only part of the story. Their success has revealed a broader opportunity: connecting with high-intent shoppers through precise, privacy-safe targeting across the open web.

That’s where commerce audiences come in.

The Ad Tech Explained Team developed this explainer, in partnership with Sovrn, to shed some light on how commerce audiences are built and the unique value they can unlock for advertisers and publishers alike.

Sponsored by Sovrn

Let’s lay the groundwork: What are commerce audiences? As in, what differentiates them from any other audience segment that advertisers might be targeting?

Commerce audiences are built from real shopping behavior, not just demographic data or broad interest categories. They’re informed by signals such as product searches, category engagement, and purchase activity, which reveal where a shopper is in their buying journey. This makes them more accurate and actionable than traditional audience segments that rely on assumptions about intent.

By combining browsing activity, product interactions, and transaction data, commerce audiences allow advertisers to target based on what a shopper is actually doing in the moment. This is especially valuable in an environment shaped by privacy regulations and the decline of third-party cookies, where durable, consented data is at a premium.

This sounds like an area where the major retail players are going to dominate. Is that the reality of it? 

Large retail media networks have undeniable value, but they operate within walled gardens that can limit reach and flexibility. The bigger opportunity is to go beyond those walls by tapping into commerce data available across the open web — from publishers, affiliate networks, and other content sources — and activating it across channels.

With the right partnerships, advertisers can use commerce audiences to reach high-intent shoppers wherever they are, whether that’s on CTV, mobile, display, or social. That means campaigns aren’t locked into a single retailer’s environment and can follow shoppers across their full journey.

OK. Let’s talk about where this fits into an advertiser’s planning. What are the distinctive benefits of commerce audiences?

Commerce audiences excel in the mid-funnel, which is that critical stage between awareness and purchase where intent takes shape and decisions are made. They help advertisers connect with shoppers who have moved beyond casual browsing but haven’t yet bought.

For advertisers, four benefits stand out when it comes to commerce audiences:

• Better audience insights: When you connect browsing behavior with intent signals and verified purchase history, advertisers get a much clearer, more nuanced view of shopping behavior. Instead of lumping people into broad categories like “electronics enthusiasts” or “fashion shoppers,” you can pinpoint the ones who are actively comparing specific models of headphones or filling online carts with dresses in a certain price range. That level of precision allows for sharper targeting and more relevant creative.

• Privacy-safe first-party data: Commerce audiences are built on durable first- and zero-party data—that is, information collected directly from consumer interactions with consent. These signals comply with privacy regulations while still delivering the accuracy advertisers need. In a world where third-party cookies are disappearing and device identifiers are harder to come by, this kind of data is becoming the foundation of responsible targeting.

• Performance alignment: High-intent audience segments are a natural fit for ROAS-driven strategies because they zero in on people most likely to convert. Instead of spending budget on large, unfocused segments, advertisers can direct spend toward those who are showing clear buying signals right now. This makes it easier to hit performance benchmarks, improve conversion rates, and prove ROI.

• Scarce purchase intent signals: Genuine, real-time purchase data is one of the most valuable (and limited) assets in advertising. Most targeting approaches rely on proxies for intent, like content consumption or demographic patterns, which can be outdated or misleading. Commerce audiences give advertisers direct access to signals that show when someone is actively moving from interest to purchase.

This all sounds like a no-brainer for advertisers. But what about publishers? Are they getting left out when it comes to commerce audiences?

Not all commerce audience strategies benefit both sides equally. In some cases, publishers contribute the data that fuels the segments without sharing in the resulting demand.

Sovrn’s approach is different. Its commerce audiences are built directly from activity on its publisher partners’ sites, so publishers retain control over their data and benefit from increased advertiser interest in their inventory. This model ensures that as advertiser demand for high-intent segments grows, publishers see more value flowing back to them.

Interesting. It sounds like Sovrn sits in a unique place when it comes to building commerce audiences. What more can you tell us?

Sovrn’s Commerce Audiences product combines two proprietary strengths: its affiliate platform, used by more than 5,000 publishers and creators, and the Sovrn Exchange, which reaches over 500 million daily active consumers. This pairing allows Sovrn to capture purchase behavior in real time, connect it with contextual and engagement data, and package it into actionable audience segments.

Because Sovrn’s data is directly sourced, it avoids the delays, degradation, and added costs that come with intermediated data. Advertisers get fresh, high-resolution intent signals, and publishers gain a channel for monetizing their commerce data without sacrificing control.

Sovrn’s scale makes this possible. Its affiliate tools power over $3 billion in annual retail spend across major categories like consumer electronics, home and garden, and fashion. The network captures more than 3 billion daily visits from search-driven traffic, creating a vast stream of purchase signals.

Early market adoption has been strong, with launch partners including Simpli.fi, Basis, and tvScientific. These companies are already activating Sovrn’s commerce audiences across campaigns that need precise targeting of in-market shoppers outside the major retailer ecosystems.

About Sovrn

Sovrn empowers publishers and creators with advertising, affiliate marketing, and data products to help them understand, operate, and grow their businesses. Serving thousands of customers across 80,000+ websites, mobile apps, and CTV channels, Sovrn reaches over 500 million active consumers daily. Sovrn is committed to transparency and fighting fraud, as recognized by IAB, JICWEBS, and TAG. Based in Boulder, Colorado, with offices in New York and London, Sovrn is a trusted leader in publisher technology. Visit sovrn.com to learn more.

Sponsored by Sovrn

Sponsored by Nexxen

You cannot target what you cannot see. Viewers are everywhere—bouncing among linear, streaming, and digital screens—which makes it harder than ever to know what is actually being watched. That is where automated content recognition (ACR) data comes in. It gives advertisers a real-time signal of what content appeared on a connected TV, offering a clearer picture of exposure and helping power smarter targeting, frequency management, and cross-screen measurement. 

The Ad Tech Explained Team developed this explainer, in partnership with Nexxen, to help you unpack what ACR data is, how it works, and why it is reshaping how advertisers think about targeting and measurement.

Sponsored by Nexxen

First things first: What is automated content recognition (ACR)?

ACR is a technology built into smart TVs that can, with a viewer’s opt-in, detect and identify what is playing on the screen in real time, whether it is a live sports broadcast, an on-demand show, a commercial, or even a video game.

So then, what is ACR data?  

ACR data is the anonymized, aggregated output from smart TVs equipped with ACR technology, collected passively and in an opted-in capacity, without relying on cookies. It tells advertisers what content people have seen and when they saw it. This data comes from millions of opted-in devices and provides a real-world signal of exposure. The ACR data provides a window into real-world viewing behavior, helping advertisers understand what audiences have seen so they can plan, target, and measure more effectively.

ACR data is all about what is happening on people’s TVs. But is it useful beyond the big screen?

Absolutely. ACR data plays a central role in omnichannel strategies by connecting TV exposure to activity across other devices.

Take Nexxen’s use of ACR data as an example. With Nexxen’s cross-screen capabilities, advertisers can:

• Extend reach across mobile, desktop, and social

• Understand how TV ads influence digital behavior

• Manage frequency across channels

• Measure the full customer journey from impression to outcome

That makes sense. How is this data specifically being put to use by advertisers?

Advertisers use ACR data to make smarter, more impactful media decisions. Some common strategies include:

• Retargeting viewers who saw a brand’s ad on linear TV with follow-ups across digital TV and other devices

• Suppressing already-reached households to reduce waste

• Engaging viewers of competitors’ ads to win share of voice

• Aligning with premium content and tentpole events to boost relevance and recall

Financial services brand Empower, for example, worked with Nexxen to reconnect with audiences who had seen their ads—and competitors’ ads—during programming for one of the most-viewed sports events. The result was an 11% lift in aided awareness, a 58% increase in brand recall, and a 50% boost in consideration.

In Australia, health insurer Bupa used Nexxen’s ACR data to exclude more than 830,000 households already exposed to its linear ads. The campaign then reached 1.2 million additional households across digital TV, driving a 16% lift in incremental reach and uncovering that 94% of FAST audiences were entirely new.

Are there any limitations advertisers should keep in mind with ACR data?

Like any data source, ACR is most effective when it is leveraged as part of a larger strategy. It reflects viewership from opted-in smart TVs. So while incredibly valuable, it should be considered one signal among many in a cross-screen world.

That’s why Nexxen is expanding its partnership with VIDAA, the operating system behind millions of Hisense and Toshiba smart TVs, which provides Nexxen with exclusive access to VIDAA’s ACR data for use by advertisers on our platform. The recently renewed agreement will grow the availability of verified viewership data on the open internet and enable more precise audience targeting and measurement. For advertisers, this means greater control over ad frequency, fewer wasted impressions from repeat exposures, and a clearer view of how TV exposure drives results across different screens.

About Nexxen 

Nexxen empowers advertisers, agencies, publishers, and broadcasters around the world to utilize data and advanced TV in the ways that are most meaningful to them. The company’s flexible and unified technology stack comprises a demand-side platform (DSP) and supply-side platform (SSP), with the Nexxen Data Platform at its core. With streaming in its DNA, Nexxen’s robust capabilities span discovery, planning, activation, monetization, measurement, and optimization – available individually or in combination – all designed to enable partners to achieve their goals, no matter how far-reaching or hyper niche they may be.

Nexxen is headquartered in Israel and maintains offices throughout the United States, Canada, Europe, and Asia-Pacific, and is traded on the Nasdaq (NEXN). For more information, visit www.nexxen.com.

Sponsored by Nexxen

Cloudflare recently announced a new feature that blocks AI crawling bots on behalf of its customers. If you're unfamiliar with a crawler, it's essentially software that analyzes web pages and categorizes the information it finds.

In the age of search, crawlers helped search engines like Google develop their massive database used to return relevant search results. Now, generative AI chat tools need to do the same thing to return appropriate responses to users. 

Blocking Bots

Publishers have always had the option to block Google crawlers or other bots, but this is akin to financial suicide for any publishers that rely on advertising revenue. Nobody with advertising as their primary business model would ever block a Google bot since it would limit the chances of your pages surfacing in search results, which bring traffic and eyeballs to your ads. 

In the age of AI, when users receive the information they need directly in a response, whether that be a Google AI Overview, ChatGPT, Perplexity, or another AI tool, they have less reason to visit a site with the same information. Publishers can instruct bots not to crawl their website by updating the "robots.txt" file on their root domain and disallowing the bots' user agent. 

If you check out the New York Times robots.txt file, you can see entries to block both ChatGPT and Anthropic in this file:

User-agent: ChatGPT-User
Disallow: /

User-agent: anthropic-ai
Disallow: /

These blocks are not surprising given the New York Times copyright infringement lawsuit against OpenAI. 

Companies are not required to obey robots.txt and can choose to ignore it; the file itself doesn't technically block a bot. But tools like those offered by Cloudflare do technically block the bot before it can hit a web server hosting content. 

Unsurprisingly, some AI scrapers may be acting nefariously to circumvent restrictions. It’s rumored that some players (Perplexity, for example) might be spoofing user agents, which they could be doing to get around blocking mechanisms. Generative AI search tools are in fierce competition, and they may seek any means to gain an edge, despite ethical or legal concerns. 

But Cloudflare recognizes these shady practices and uses advanced analysis to detect spoofing and other evasion mechanisms. The company now asks all new clients if they wish to block any bot it finds (while providing the same tools to existing customers). 

Additionally, Cloudflare is looking to create or facilitate a "pay per crawl“ market that would compensate content creators for the content they produce by giving publishers the capability to signal to AI agents that they have to pay to crawl a site. 

Cloudflare wants to do this using existing HTTP status codes:

So Cloudflare could allow or deny access to content using HTTP status codes, and act as the broker to collect payment:

It's a cool idea and an admirable pursuit, but it also feels like a last-ditch effort by a company that is sure to see a monetary impact if the web contracts in any way due to usage-based pricing. After all, Cloudflare offers web infrastructure services to site owners, so if fewer people are navigating the web, then there is less need for Cloudflare's services. Much like ad tech companies and publishers that rely on open web advertising, the rise of generative AI chat and search directly threatens Cloudflare's business.

Meanwhile, Perplexity offered a glimpse of how a different (more closed) model to compensate publishers could work. As you might imagine, there are some notable differences between Cloudflare’s and Perplexity’s approaches: 

• The Cloudflare model compensates publishers for any crawl for AI training, whereas Perplexity only shares revenue when it uses specific publisher content to formulate a generative AI response. 

• Additionally, Cloudflare's approach aims to establish an open standard for all companies to adopt, whereas Perplexity's model is a direct arrangement between a publisher and Perplexity.

In either case, we are inexorably marching toward a future where AI relegates any content creator to a mere input mechanism for insatiable AI models. Google once controlled the front door to most content, but now that entryway is closing in favor of an always-on digital information siphon.

Each publisher that adopts Cloudflare's pay-per-crawl model, inks an AI content licensing deal, or joins Perplexity's revenue-sharing program is attempting to extract value from a seismic shift occurring in the economics of open web content creation. However, they might also be signing their own death certificates, as they give more credence to the very models that will obviate the need to visit any publisher's property.

The future of the web

So what's to become of content creation on the web? If you project forward trends that are already happening, I think this is where you end up:

1. Web traffic is declining due to generative AI chat and search.

2. Web publishers are losing advertising revenue due to fewer visits to their properties.

3. Advertising revenue is shifting to destinations not impacted by this phenomenon. 

4. Remaining web publishers and content publishers are pivoting their business models to capitalize on these trends.

I don't believe that all web publishing will die, but a majority will. Gone are websites that rely on search traffic, such as SEO parasites that cater to simple queries like "where can I stream tonight's NFL game?" or recipe sites. 

Would you rather scroll this ad-riddled hellscape to find a chicken marinade or have it presented to you in perfect formatting with the ability to ask follow-up questions? The answer is clear. 

Ad-riddled hellscape or beautifully formatted perfection? You choose.

Cloudflare's pay-per-crawl model reminds me of the doomed micro-transaction model, where users would pay a minimal fee to access content. But everyone quickly realized that paying for content from unknown entities made little to no sense, as users were too accustomed to free access to general content. Users, however, were willing to pay media brands and individual content creators they trust.

Twitch proved this model in video for individual creators, and it has since transitioned to the written word with platforms like Substack. Investors may understand that unique human thoughts amid a sea of AI-generated content will stand out and hold value in the future, as Substack recently raised $100 million at a $1.1 billion valuation. 

Forming a direct connection with users via subscriptions remains a sustainable business model in our AI-driven future. And, perhaps it's not all doom and gloom for advertising on the web.

If the web contracts and web advertising supply along with it, then naturally, one would expect prices to rise on the remaining inventory. If publishers can endure the initial reduction in traffic and adopt alternative business models (such as subscriptions or AI content licensing), then the remaining web inventory might see some price improvement due to scarcity.

The more likely scenario is that we will see an acceleration of advertising spend flowing to alternative sources. The bad news for most publishers is that the most likely landing spot for this revenue is the very tools displacing the entire web, along with the social platforms pumping out AI-generated content that will keep the masses entertained in a dopamine-fueled endless scroll of blissful satisfaction. 

The good news for streaming publishers is that generative AI is less likely to displace video content in the short term, and could make this inventory more valuable as supply from the web dries up. CTV becomes increasingly valuable as advertisers seek to expand their reach beyond the behemoth platforms, with nowhere else to turn.

I weep for the future (humans)

Despite the best efforts of companies like Cloudflare, it's hard not to be pessimistic about the future of web publishing that relies on advertising.

As open web advertising economics decline, more content will shift to walled gardens. In these gardens, content creators might be required to allow the use of their content to train models (e.g., Gemini in the case of Google, or Llama in the case of Meta) if they want to publish on YouTube or Instagram. 

To take the most pessimistic view possible, most content creation will likely move beyond human control. Impossibly beautiful influencers (because they aren't real) are gaining follower counts rapidly. Google's NotebookLM generates podcasts on demand about any topic. And virtual content creators are now emerging in YouTube videos. 

There will always be a craving for authentic, artisanal-crafted human content, but it could eventually represent the vast minority of all content on the internet. Cultural backlash against AI-generated content is real, but I think most humans will gravitate toward any content that entertains or delights them.

Ads.txt has helped protect the programmatic ecosystem from fraud and bad actors for nearly eight years now. The IAB Tech Lab published the original Ads.txt specification in 2017. Since then, it has enabled publishers to declare who is authorized to sell their inventory, providing DSPs with a tool to help eliminate counterfeit inventory from their platforms.

So, why am I writing an article about Ads.txt in 2025? When it comes to ads.txt, I thought it was a simple concept I could learn once, implement, and then forget about. However, the added complexity of CTV and, more recently, the growing sophistication of DSP enforcement have compelled me to take a closer look at how the specification has evolved.

Initially developed for the web only, ads.txt has undergone modifications over the years to keep up with behavioral trends shifting from the web to apps, mobile apps to CTV apps, and the business trend of syndicating video content outside a publisher's owned and operated apps. 

Publishers, buyers, or ad tech vendors who want to understand these developments and how they apply to CTV would have to pore through the various iterations of the Ads.txt spec and associated documents released by the IAB Tech Lab. To write this, I had to slog through:

• Ads.txt v1.1

• App-ads.txt v1.0

• Ads.txt implementation guide

• ORTB 2.6

• ORTB SupplyChain Object GitHub

• IAB Tech Lab OTT/CTV Store Assigned App Identification Guidelines

Ads.txt and its menagerie of complimentary standards has quietly evolved into one of the more complex concepts that anybody working in programmatic should understand. The IAB Tech Lab has woven a wicked web of documentation to untangle and has successfully produced the most dry reading that I’ve ever had the displeasure of laying my eyes on. 

Fortunately for you, I’ve developed a complete immunity to dry ad tech documentation and created this guide to help you quickly understand how ads.txt applies to CTV. 

While this explainer cannot serve as a replacement for anyone implementing these technical standards (sorry, you have to read the documentation), I hope it will provide anyone who wants to learn with a solid understanding of how publishers and programmatic monetization managers set up ads.txt for CTV and how DSPs validate supply paths.

One Minute Ads.txt Refresher

We need to cover some basics before delving into more advanced CTV and ads.txt concepts. Here is your one-minute ads.txt refresher.

DSPs use Ads.txt to combat domain spoofing or sellers misrepresenting the origin of inventory. Sell-side platforms often offer tools to manipulate the values sent in a bid request, such as a web domain (example.com) or bundle ID (essentially a domain for web like com.example.androidTV).

Publishers do have valid reasons to manipulate these values (much to the annoyance of advertisers and DSPs). But bad actors also manipulate bid request values to either conceal the true origin of the inventory or inflate the value by pretending the request originated from a premium property.

Ads.txt enables publishers to declare who is authorized to sell their inventory publicly. DSPs then cross-reference these lists and have the option to reject bid requests from unauthorized sellers.

Let’s quickly look at a typical ads.txt workflow on the web:

1. SSP sends a bid request for a publisher that contains a domain value and publisher.id value. Ex: publisher.com and 123456

2. DSP takes the domain value and appends “/ads.txt” and checks that file. Ex. publisher.com/ads.txt

3. DSP enforces internal or client rules based on ads.txt values.

In the ads.txt files, publishers list the SSPs they work with, along with their account IDs. 

Example:

pubmatic.com, 123456, DIRECT

If you see this line in a publisher’s ads.txt file, you know they work with Pubmatic, their account ID is 123456, and the publisher directly controls any inventory originating from that account.

Here is an example of a very neatly organized ads.txt file: https://www.accuweather.com/ads.txt.

AccuWeather even goes a step further, listing out each DSP they are working with by SSP (which is not a requirement). Notice the #? This symbol instructs DSP to disregard anything that follows it; it’s essentially internal commentary for Accuweather’s record-keeping purposes.

What are DSPs checking for, you ask? DSPs can either choose to apply platform-level ads.txt requirements or leave up enforcement to their clients through an account or deal targeting features. Some DSPs require a valid ads.txt, while others can leave that decision up to the clients. By “valid ads.txt” I mean:

1. The ads.txt file exists (based on the domain sent in the bid request)

2. The ads.txt file found on the domain from the bid request lists the account ID from the SSP initiating that same bid request.

3. (Optional) The ads.txt listing is DIRECT (if the client or DSP chooses only to buy non-resold inventory)

Furthermore, they can check an SSP’s sellers.json to confirm the company behind the bid request.

Sellers.json is a file hosted by SSPs that lists all entities with accounts on the platform. Each entry in the file lists the entity's company name, account ID, company website, and what type of seller they are. This file enables the DSP to identify the exact company associated with an account and perform additional validations.

An example line in sellers.json for a publisher looks like this:

{

"seller_id": "123456",

"name": "Totally Legit Publisher LLC",

"domain": "publisher.com",

"seller_type": "PUBLISHER"

}

OR 

{

"seller_id": "78910",

"name": "Stinky Reseller Inc.",

"domain": "ilovearbitrage.com",

"seller_type": "INTERMEDIARY"

}

DSPs can use the additional information in the sellers.json file to further confirm whether it’s a direct publisher’s account initiating a bid request or an intermediary by validating the “seller_type” field.

DSPs can choose to use this information in further enforcement (e.g., not allowing any intermediaries). SSPs can also list accounts as “BOTH” – which some DSPs may frown upon, instead preferring to keep direct and reseller business on separate accounts for clarity.

Example sellers.json: http://pubmatic.com/sellers.json

Ok, maybe that was more than one minute, but trust me, I saved you at least an hour and a headache. 

CTV app-ads.txt

CTV leverages the app-ads.txt specification, which changes a few critical elements above. 

First, instead of using a domain value to find the ads.txt file for verification, they use an app store URL sent in a bid request to find a path to an app-ads.txt file. 

Example: https://apps.apple.com/us/app/tubi-movies-live-tv/id886445756

Once a DSP has an app store listing, they then need to look for a developer URL. They will use this URL to find the publisher’s app-ads.txt file by appending “/app-ads.txt” to the URL they find. 

This process should be easy if app stores follow the IAB Tech Lab guidelines and insert a piece of metadata into the HTML of an app store listing web page:

<meta name="appstore:developer_url" content="https://www.developerurl.com" />

This meta tag is a piece of metadata embedded in the HTML of a web page that is invisible to a typical web user but accessible to external systems interested in the information.

But if you navigate to those Apple app store examples above and view the page source, you’ll notice that meta tag is nowhere to be found in the web page HTML. Apple doesn’t do a small little thing to help make the programmatic advertising ecosystem a little better. Shocker!

Maybe Roku will be a little more accommodating. If we navigate to the Roku channel listing for Disney+:

https://channelstore.roku.com/details/8f8e53a488990d6da45c79e914ed3785:33140619236ca42d45a9e4b4eecb5c41/disney-plus

We see that the very advertising-friendly company Roku added the required metadata to validate app-ads.txt:

<meta name="appstore:developer_url" content="https://disneyplus.com"/>

Any DSP validating Disney+ bid requests can then use the URL on the Roku channel listing page to build the URL to the Disney app-ads.txt file: disneyplus.com/app-ads.txt. From this point, a DSP can perform all the necessary validation as described in the prior section. 

This workflow is ideal if the CTV platform supports public app store listings for streaming apps; however, this is not always the case. For example, there are no public store listings for streaming apps on the Sony Playstation store — so DSPs cannot validate inventory originating from this platform.

As for the Apple App Store listing, DSPs would need to build a custom web page scraper to extract the developer URL from the "Developer Website" link on App Store pages.

SupplyChain Object

The SupplyChain object is a fun little diddy added in ORTB 2.6 or as an extension in ORTB 2.5. It allows DSPs to understand the provenance of any opportunity sent in a bid request and understand all the entities involved in a given transaction.

The SupplyChain object provides the DSP with insight into the number of hops between platforms or companies a transaction took, which they can use to eliminate multi-hop inventory or to understand exclusive reseller paths (where the publisher outsources monetization to another company).

DSPs can find the supply chain in the “source” object in an ORTB bid request. Here’s a multi-hop example with a reseller (resellerssp.com) and a publisher (publisherssp.com). 

"source": {
    "ext": {
      "schain": {
        "ver": "1.0",
        "complete": 1,
        "nodes": [
          {
            "asi": "publisherssp.com",
            "sid": "12345",
            "hp": 1
          },
          {
            "asi": "resellerssp.com",
            "sid": "6789",
            "hp": 1
          }
        ]
      }
    }
  }

The DSP can see a multi-hop sale and choose to ignore it, focusing instead on direct inventory, or they can also use this in conjunction with the two other ads.txt values below to gain a deeper understanding of a given transaction.

OWNERDOMAIN + MANAGERDOMAIN

These two values were incorporated into the ads.txt spec to establish a stronger connection between sellers.json files and to express how some multi-hop supply chain objects represent the most direct path to inventory.

OWNERDOMAIN is the website of the corporate entity that owns a website or app. Often, companies own many websites and apps, which may not always be tied back to a sellers.json file.

In an SSP’s sellers.json file, a publisher’s listing will typically only list the corporate entity itself, not all the websites or apps they own. So by placing an “OWNERDOMAIN” value in ads.txt files, it can allow a DSP to confirm the company that owns the property and link it back to the entity sending bid requests from an SSP via that SSP’s sellers.json file. 

MANAGERDOMAIN is another value a publisher can place in their ads.txt file to broadcast to the world that they have outsourced programmatic monetization to a separate company. They are informing everyone who wants to verify the ads.txt file that they have authorized a separate company to send bid requests for inventory on the publisher’s property. 

Publishers authorizing a MANAGERDOMAIN allow a DSP to confirm that it is the most direct path to the inventory despite the manager being a reseller. Let’s look at a multi-hop supply chain example that represents the most direct path to inventory:

A DSP receives a bid request from account ID 1234 (manager.com) from an SSP (ssp.com) for an opportunity on a CTV app. The bid request contains the following supply chain:

"source": {
    "ext": {
      "schain": {
        "ver": "1.0",
        "complete": 1,
        "nodes": [
          {
            "asi": "manager.com",
            "sid": "5678",
            "hp": 1
          },
          {
            "asi": "ssp.com",
            "sid": "1234",
            "hp": 1
          }
        ]
      }
    }
  }

After inspecting the app store URL of the CTV app from the bid request, the DSP extracts a developer URL of “ctvapp.com” and visits “ctvapp.com/app-ads.txt.” The DSP then finds these contents in the ads.txt file:

OWNERDOMAIN = publisher.com
MANAGERDOMAIN = manager.com

manager.com, 5678, DIRECT
ssp.com, 1234, RESELLER

From this, the DSP can ascertain that “publisher.com” owns the CTV app and is outsourcing programmatic monetization to manager.com. The DSP can then visit the sellers.json files of both the Manager (manager.com/sellers.json) and the SSP who sent the original bid request (ssp.com/sellers.json).

Even though the manager isn’t an SSP themselves, this situation calls for them to host a sellers.json file to allow DSPs to verify the corporate entities with which they do business and create a linkage between the publisher’s ads.txt file. 

manager.com/sellers.json:

{

"seller_id": "5678",

"name": "Publisher LLC",

"domain": "publisher.com",

"seller_type": "Publisher"

}

ssp.com/sellers.json:

{

"seller_id": "1234",

"name": "Manager Inc.",

"domain": "manager.com",

"seller_type": "Intermediary"

}

A DSP will now know that:

1. Manager Inc. is the seller initiating the bid request (seller ID 1234 on ssp.com/sellers.json = Manager Inc.)

2. Manager Inc. is the valid manager of Publisher LLC inventory, as ssp.com/sellers.json domain for seller ID 1234 = MANAGERDOMAIN listed in ctvapp.com/app-ads.txt (manager.com).

3. Publisher LLC (publisher.com) has an authorized and direct relationship with Manager Inc. (manager.com) since Publisher LLC listed MANAGERDOMAIN = manager.com on ctvapp.com/app-ads.txt, and there is a DIRECT entry matching the seller ID in the supply chain object on manager.com/sellers.json.

As I mentioned at the beginning of the article, this is a complex process. Hopefully, you are still with me because we have one last ads.txt value to cover. 

Inventorypartnerdomain

CTV streamers run owned and operated apps, but they can also license their content to other apps, such as vMVPDs (e.g., Pluto, Tubi, Sling, etc.). These deals typically include inventory shares, where the vMVPDs will send ad requests to a content owner’s ad system, allowing them to monetize the opportunity. 

In this scenario, the opportunity is technically "direct," as the content owner owns the opportunity, as defined in an inventory share agreement. The vMVPD would send an ad request to the content owner’s SSP, which then initiates a bid request to a DSP.

However, things can get messy since the DSP will check the app store listing in the bid request, which will be the listing of a vMVPD app, not the publisher’s app. The DSP would then find the vMVPD's developer URL and check the app’s app-ads.txt file hosted there. 

The vMVPD would then be responsible for listing all of the content owner’s SSP seats and account IDs and marking them as 'DIRECT' to indicate that this inventory is owned directly by the content owner. This workflow is less than ideal for a couple of reasons:

1. It creates confusion since the account IDs in each line in the vMVPDs app-ads.txt file would be a smattering of all the various content owners they license content from and have inventory sharing agreements with (along with their own).  

2. It requires the vMVPD to update their app-ads.txt file every time a content owner changes or adds a new ad system.

Inventorypartnerdomain solves this challenge. The vMVPD can instead list a single line that points to the content owner’s app-ads.txt file.

Example:

inventorypartnerdomain=contentowner.com

The content owner will also pass this value in the site or app object of the bid request they initiate for inventory on the vMVPD app:

"app": {
    "bundle": "1234",
    "storeurl": "https://appstore.com/details/vMVPDapp/1234/",
    "inventorypartnerdomain": "contentowner.com",
    "publisher": {
      "id": "5678"
    }
  }

Now, the DSP first checks the vMVPD’s app-ads.txt file, which should contain a line listing “inventorypartnerdomain=contentowner.com.” Then they can proceed with checking contentowner.com/app-ads.txt, and there they should find:

ssp.com, 5678, DIRECT

That authorizes the original bid request sent from ssp.com and account = 5678. The content owner can now handle all ads.txt updates in their file independently of the vMVPD. 

Congratulations, you now have a rudimentary understanding of how ads.txt works in the context of CTV. As you may have noticed, creating a valid and verifiable CTV supply path is a complex process that involves many steps and the coordination of several parties. However, the result is a much safer programmatic ecosystem with fewer bad actors and illicit supply chains.

Sponsored by PayPal Ads

For more than two decades, PayPal has been building a digital payments brand that has become synonymous with secure transactions and seamless checkout experiences. With a presence in nearly every corner of e-commerce, the company has established deep connections with both consumers and merchants, giving it a unique vantage point on the evolving world of digital commerce. In October 2024, that vantage point became something more: a launchpad for PayPal Ads, the company’s ambitious new foray into advertising. With this launch, PayPal is leveraging its role at the intersection of payments and intent to redefine what advertising can look like in a commerce-first world. 

The Ad Tech Explained Team developed this explainer, in partnership with PayPal, to help you understand the unique value proposition at the heart of PayPal Ads, with a look at what to expect to see next from PayPal Ads, especially as the industry turns its attention to Cannes.

PayPal is a name most people associate with payments, not media. What changed to make now the right time for PayPal to become a leader in the advertising business?

We’re entering an era of what we call the empowered shopper. Consumers today are in control of their shopping journey unlike ever before. They increasingly expect personalized offers, agentic experiences, and zero-click commerce – whether they’re interacting with a brand on a website, through a chatbot, or via an AI agent. Before consumers went to the store. Now, the stores must come to them. This requires a new kind of advertising infrastructure to support open commerce.

At the same time, targeting based on things like cookies is breaking down. Brands need a better way to anticipate intent and measure outcomes. That’s why I like the phrase payments are the new cookies. Purchase data is accurate and verified. It reflects real consumer behavior, not proxies. But consumers want trust, so if this can be designed with privacy at its core, it offers the best signal for understanding who to reach, when, and how.

The key here is that PayPal is a two-sided network that can create a flywheel benefitting both brands and consumers. For brands, PayPal sees billions of shopping moments a month across tens of millions of merchants and hundreds of millions of consumers — all within a trusted, privacy-first environment. This PayPal Transaction Graph provides cross-merchant intelligence to grow market share, which only PayPal can offer as a leader in commerce for around 25 years. For consumers, we can create more relevant, personalized, seamless shopping experiences that drive more conversion and engagement.

What does commerce media mean in the context of PayPal Ads — and how is it different from traditional financial or retail media? 

Retail media networks are typically limited to a single retailer’s ecosystem. They see what happens in one store, and that’s where the ad opportunity and intelligence begin and end.

PayPal, on the other hand, is connected to commerce across the internet. We see transactions across thousands of categories and tens of millions of merchants. That gives us a much more complete view of consumer behavior. It allows us to help advertisers plan, personalize, and optimize campaigns across a broader ecosystem with privacy in mind.

This starts with exciting new ad experiences on PayPal, Venmo, and PayPal Honey. It expands to PayPal Offsite Ads, where we use the PayPal Transaction Graph to improve advertising across the open commerce. And in the future, we’ll actually bring the storefront to the consumer so they can buy products with PayPal and Venmo without ever leaving the ad itself.

You sit on a massive amount of consumer behavior data. What kinds of signals are most valuable to advertisers, and how do you ensure they’re used responsibly?

The foundation is PayPal’s payments network, allowing for cross-merchant scale and insights. Approximately 70% of Americans have a PayPal account, and around half access PayPal several times weekly. Venmo gets referenced every nine seconds across social platforms, with about half of Venmo users being Gen Z or Millennial. This all occurs during a crucial consumer moment – when individuals are considering purchases and making spending decisions.

Advertisers have been most excited about understanding what shoppers buy before they come to their store, how shoppers are using text and emojis in peer-to-peer payments, and how they cannot just drive incrementality but actually grow category market share.

An important point here is that we use these signals responsibly. Privacy and trust are foundational to PayPal’s business, and that extends to how we build PayPal Ads. We’re interested in helping them discover relevant products in the right moments in ways they can trust.

PayPal reaches consumers in some very specific moments — during checkout, payment, even when checking account activity. How do you think about aligning ad experiences with those moments?

Intent is vitally important to a successful ad product. What makes PayPal unique is that we’re present during real shopping activity, not just when someone is scrolling or searching, but when they’re making a purchase, checking an order, or managing their money. That gives us an opportunity to make advertising useful and smart, rather than interruptive.

We think about how to enhance those moments, not distract from them. For example, after someone checks out with PayPal, we might show them a personalized offer based on their recent purchases. Or if someone is tracking an order in their app, we can surface a complementary product or offer from a trusted brand. These are high-intent, high-attention moments that can add value for everyone involved.

What types of advertisers or campaign goals are best suited for PayPal Ads today — and how might that change as the platform grows?

Right now, we’re seeing excitement from advertisers that are endemic to PayPal – think retail and travel, and more. These brands are focused on quality reach to high-value audiences both on PayPal properties and beyond. They want to measure attribution and market share using our transaction data.

As the platform grows, we expect our audience solutions to expand to non-endemic brands, new ad units, SMBs, and more. Our long-term vision is to support the full advertising lifecycle, from awareness to loyalty, all rooted in real commercial behavior.

What should we expect to see next from PayPal Ads — especially as the industry turns its attention to Cannes?

Cannes is exciting not only because it’s the largest ad event of the year and the only place that really celebrates the creative work of our industry. It’s also exciting for us because this year is the first time PayPal will be showing up. We’ll be powering Influential Beach and have some exciting announcements in store around shoppability, commerce, and partnerships. It’s an energizing time to be building ads that pay off.

About PayPal 

For over 25 years, PayPal has been at the forefront of the digital commerce revolution. Creating innovative experiences that make moving money, selling, and shopping simple, personalized, and secure, PayPal empowers consumers and businesses in approximately 200 markets to join and thrive in the global economy. For more information, visit www.paypal.com.

Magnite's acquisition of SpringServe back in July 2021 felt like a natural addition to the company. SpringServe offered a video ad server that could complement Magnite's robust programmatic video offerings.

Rubicon Project had immediately positioned itself as a leader in programmatic video after a merger with Telaria (and the subsequent birth of Magnite) and the acquisition of SpotX. By “Hoovering” up SpringServe (the only realistic option outside GAM and FreeWheel in a publisher's consideration set for an ad server), the company collected the final infinity stone in the Magnite video gauntlet. 

Since Magnite already had the video SSP covered, owning the other key piece of technology that every video publisher needs, an ad server, would allow them to not only make customers stickier but also allow them to optimize the entire ad serving workflow, whether it be direct, programmatic, or combining the two. 

But as anyone who has gone through a merger or acquisition before knows, buying the technology is the easy part — seamlessly integrating the infrastructure and operational processes of disparate technology platforms is where the rubber meets the road. 

After an acquisition, you must make hard decisions around duplicate features, processes, and people. You must merge the parts that make sense and cut the redundancies — which takes up valuable space on product roadmaps. But you can also develop new products and opportunities that weren't before possible.

Elegantly combining acquired pieces has been Magnite's journey for the last five years — first with the unveiling of ClearLine and culminating with their recent announcement of the new SpringServe platform that combines the legacy SpringServe ad server with the Magnite Streaming SSP. I've been following this journey, and Magnite recently invited me to dig into the latest with ClearLine and SpringServe and share what I learned with all of you. [editor’s note: this is not a sponsored post]

ClearLine

Going into my conversation with Magnite, my curiosity centered around the motivation to develop ClearLine in a market filled with buying tools for advertisers of all shapes and sizes. What differentiation could Magnite offer to lure buyers onto the platform? This angle was my first line of questioning for Magnite SVP of Product Paige Bilins and VP of Product Rory Edwards. 

We began with a demo since I had never seen the ClearLine platform before this meeting. Before the demo, I wondered how much ClearLine leveraged SpringServe tech. As a SpringServe user, I immediately noticed how familiar the platform looked, and Paige confirmed my assumption by asking, "Look familiar?" with a knowing smile.

I had known through the press that SpringServe technology undergirded ClearLine, but this flows through the platform's design language. While we can't "see" the backend of ClearLine, the repurposed UI demonstrates how Magnite leveraged its acquisition to streamline the launch of a new product to market. 

The team explained that one of the goals of ClearLine is to create a simplified buying platform for advertisers who don't necessarily need all the bells and whistles of a full-blown DSP.

They succeeded at that, as it was evident that even someone with a whiff of ad ops experience could launch a campaign on this platform lickety-split. I couldn't help but think of grossly utilitarian DSPs I've had the displeasure of using, and some other convoluted platforms came to mind.

However, there is no shortage of buying platforms focused on CTV that aim to simplify workflows as a differentiator, so I started drilling down into some burning questions about the reason for bringing this product to market. 

The Integrated Strategy

It's fine and dandy to build a simple way to buy streaming video, but what can ClearLine offer that other buying platforms cannot? After talking to the Magnite team, I realized this may be asking the wrong question entirely. It's not just what a buying platform can offer but what a completely integrated video platform can offer. 

Working within the programmatic video ecosystem can often be a frustrating experience — especially in a nascent environment (relative to display) like CTV. Programmatic advertising involves a delicate song and dance between various companies involved in any transaction. 

Order management systems, publisher ad servers, advertiser ad servers, verification vendors, SSPs, DSPs, SSAI stitchers, and more can all play a pivotal role in delivering a campaign. All these players and systems must agree on protocols and processes to achieve harmonious programmatic delivery. 

From years of experience and consequent frustration, orchestrating these systems can be more like cacophony than harmony. Programmatic ostensibly offers the benefits of glorious automation and efficiency — yet its participants' competing motivations and priorities often lead to a lack of innovation.

And herein lies a key motivation for Magnite with the recent integration of their ad server + SSP and launch of ClearLine: streamlining programmatic activation. 

SSPs and DSPs have independent customer sets that drive independent desires and strategies. Let's say, for example, that Magnite had a set of publishers who wanted to offer programmatic pause ads — to make this a reality, Magnite would have to 

1. Find a DSP willing to test this new type of ad unit

2. Agree on request/response protocols (ex. ORTB/VAST + Display/Native/Non-linear video)

3. Agree on creative specifications (e.g., 1920x1080 image - is transparency allowed? Is full frame allowed?)

4. Prioritize development on independent roadmaps

5. Wait until both sides finish software development

6. Find test publishers and test advertisers

7. Align testing (Maybe at different priorities at both companies)

However, since Magnite has an in-house buying platform, it has a built-in buying partner and can dictate when to launch a new ad unit on its own accord. It can align internal teams and priorities to rapidly launch a way for buyers to purchase pause ads using a self-service buying platform. 

So Magnite did just that because buying self-serve pause ads is a real feature on Magnite that is available now.

Magnite could skip steps 1-7 above and launch an innovative CTV ad unit via self-serve on ClearLine without waiting for another company to get their act together. The company can then apply what it learned internally to speed up the enablement of programmatic pause ads sourced from external platforms.

Accelerating CTV

New ad formats are cool, but what about the down-and-dirty parts of programmatic CTV? I started probing the Magnite team for more examples of how an integrated approach like this could better serve publishers and advertisers alike. 

We started getting into topics that hit close to home, as some of these issues have plagued my existence during my career. I'm going to list a few examples of how the disassociation of the SSP and DSP causes major headaches for CTV publishers and advertisers:

Fixing Programmatic Guaranteed

Programmatic guaranteed deals make up a critical component for some streaming publishers. As the name suggests, they guarantee revenue for the publishers and guarantee that publishers will fulfill advertisers' budgets in full. However, there are a hundred different ways that a PG campaign may not deliver, and a lot of it has to do with the dissociation between the supply-side and demand-side platforms. 

If the SSP and DSP aren't completely in sync around the parameters of the deal, the simplest things like a misconfigured flight date, budget, or incompatible creative can derail a PG deal in its tracks. Nailing setup on both sides is imperative with PG relative to PMP because of the constricted time frame, and publishers often reserve the inventory in their system, which impacts planning.

The stakes are even higher during live events where PG errors can blow up an entire opportunity and lead to dreaded slate (no ads and a "we'll be right back" message), which annoys users and destroys monetary potential. 

SSPs and DSPs can do their best to mitigate these issues with things like deal syncing, creative pre-ingestion, and sharing creative fail reasons via an API, but it's still not enough to ameliorate all PG issues.

Sometimes advertisers update deal settings after a deal sync between the systems, advertisers don't pay attention to failure reasons, or DSPs don't have mechanisms to communicate publisher creative specifications automatically.

This lack of proper, all-encompassing integration thrusts the promise of programmatic (automation and efficiency) back into the manual stone age. I often feel like beating my laptop with a stick like a caveman each time I have to solve these types of issues. 

A single integrated platform could save everyone the headaches that come with integrating disparate platforms. Magnite understands these woes and has already started moving in a direction with ClearLine to address them. 

Paige explained how Magnite locks ClearLine targeting for PG deals, eliminating one of the issues mentioned above. Typically, publishers should be the only side applying targeting on PG — Magnite can ensure this is the case by locking the deal settings for the buyer once the deal is activated. Even though this is a single example, Magnite now also has the flexibility to continuously improve deal and creative syncing between platforms without waiting to adopt any updates to creative or deal APIs.

Live Events

For something as advanced as the programmatic monetization of live events, some very manual processes are still necessary to ensure smooth delivery. One of these is ensuring DSPs understand the potential spike in QPS (queries per second) that a live event may cause. A deluge of bid requests could tip over a DSP's systems if they aren't prepared. 

Publishers must manually provide QPS estimates for each live event via email, another primitive process baked into a “programmatic” workflow. The industry is working on some standards to at least share these estimates between systems, but those specifications have not yet materialized, and even when they do, it will take more time for adoption. 

Magnite is working on forecasting tools that will allow publishers to preregister live events so it can not only assist with the publisher's forecasting capabilities, but it can also push those expected spikes into ClearLine to assist with scaling and pacing — eliminating the shock and risk that live event spikes can bring.

Speaking of forecasting, since ClearLine has insight into the ad server and SSP, it has complete visibility into supply and demand — an enviable position to be in relative to traditional DSPs who only see the segment of inventory the publisher chooses to share — making forecasting, pacing, and optimization much more difficult. 

Content Metadata

CTV buyers want more insight into the shows and episodes they buy, but publishers must consider laws and regulations that prevent exposing information like this in bid requests. It is generally off the table to share anything more than generic content metadata like genre and rating in bid requests containing user identifiers like device IDs, hashed email IDs, or IP addresses since this could reveal the person's viewing habits. 

There is a world where a DSP could ensure that no person or client can get their hands on both the user identifier and show/series title from the same transaction, but publishers have no technical mechanism to ensure the DSP holds up that promise. Once a publisher combines these two pieces of metadata and blasts it into a bid request, it's out of their hands. 

But what if the publisher's and buyer's technology were on the same stack? Another potential benefit is that a platform like SpringServe could share the sensitive metadata between systems and ensure that no buyers could access user identifiers + show/series names together.

Since they own the end-to-end workflow, they have complete control over handling that data which could open up show level targeting and reporting that wouldn't otherwise be possible between two separate companies on the supply and demand sides.

There are also no worries about mapping SSP genre or category values between systems that often use separate taxonomies, leading to unintended targeting or reporting results.

Pricing

The above examples are not an exhaustive list, but I wanted to move on to another topic that always came across as a potential primary benefit of ClearLine: pricing.

One of the promised benefits of ClearLine is that advertisers can devote more spend to working media. By nature of being built in the SpringServe ad server, ClearLine offers a more direct connection between publisher and buyer than buying through a DSP. And I was told that "The cost to activate purposefully reflects that level of directness."

I figured I wouldn't get many specific answers on pricing, but I did receive some insight from Magnite: 

It's a journey

Before I wrote this article, I viewed Magnite as attempting to take on traditional DSPs directly with ClearLine. After talking with Magnite and writing this article, I understand that launching any buying tool will naturally pull away budgets from other buying products, but I now view the strategy as more nuanced. Magnite wants to offer a more seamless CTV activation path for both buyers and sellers.

By integrating the ad server, SSP, and buying platform, they can remove the friction often involved when coordinating technology integration between separate companies. They hope to speed up activation time and solve previously thorny issues that external collaboration manifests.

At the same time, Magnite continues to consolidate its technology through the combination of Magnite Streaming and the SpringServe ad server. Rather than wholly separate platforms, Magnite looks to introduce ad serving and programmatic via separate but now much more interconnected modules to speed up operational workflows (and eliminate the legacy technical debt of maintaining separate platforms — which should allow them to focus more time on launching new features).

Paige told me that integrating the ad server and SSP is a "journey" and ongoing. The beta of the new integrated platform is live now, and the feedback they earn is sure to mold the direction of the technology and dictate how else the integration of ClearLine can benefit publishers and advertisers alike.

With a five-paragraph blog post, Google has officially cemented the Privacy Sandbox as the biggest waste of time in ad tech history. 

The company has decided not to add a prompt in Chrome asking users if they'd like to disable third-party cookies, which means not only will cookies not be deprecated, but everything will stay the same as before Google first announced it would deprecate cookies five years ago.

In other words, since nothing will change in Chrome, we have all wasted a good chunk of the last five years of our lives thinking, debating, and writing about the impending deprecation of third-party cookies and the Privacy Sandbox proposals meant to replace their advertising functions.

While I limited my exposure to mostly thinking and writing about the Privacy Sandbox, I weep for the poor, unfortunate souls and companies that dedicated actual capital, resources, and their product roadmaps toward the doomed endeavor. It makes my head spin thinking about the millions of dollars and countless hours wasted on the Privacy Sandbox. 

Writing about the Privacy Sandbox dominated the early days of Ad Tech Explained. It provided for good content at the time, but wasted brain space in retrospect.

Whenever someone asked for my honest opinion about the Privacy Sandbox and if they should test it out, my answer remained unchanged since I wrote my first Sandbox article in 2020: don't do anything. I received bewildered looks from colleagues and peers but remained steadfast in my opinion.

I knew a handful of companies were taking the immense risk of testing the menagerie of absurd APIs for the potential payoff of positioning themselves ahead of everyone else should Google pull the cookie plug. If they found the APIs could achieve parity, then my thinking and advice would be to take the late-mover advantage of avoiding all the early pitfalls (but without the added benefit of being ahead of the game). Unless you wanted to risk the massive distraction and dedication of resources for a potential payoff that came with an early jump, a wait-and-see approach always seemed like the most prudent decision.

As I learned and wrote more about each ridiculously complex Privacy Sandbox API proposal and the building momentum of antitrust scrutiny against Google, my opinion only hardened. Introducing the esoteric Sandbox proposals into the digital advertising ecosystem always seemed like a ludicrous prospect to me, given that there are still so many troubleshooting challenges and problems that still need solving within the current state of ad tech.

Additionally, once the antitrust scrutiny upon Google switched from a slow trickle to a full-on deluge, I knew this whacky side project was standing on shaky ground. 

The fact that the capricious strategic whims of a single company could reshape the core workings and dictate the product development strategy of an entire industry should have given everyone pause from the get-go. A company holding this much power to steer every company's strategy in ad tech screams monopolistic issues. So it's no surprise that this final decision by Google has landed right as the courts have ruled Google is an illegal monopoly twice over: in search and ad tech.

The main issue with the decision to remove cookies has always been Google performing a balancing act between privacy scrutiny and antitrust concerns.

Removing cookies would have further entrenched Google's dominance in digital advertising by making web advertising less effective for everyone except Google, which sits on mountains of valuable first-party data, a monopoly search business, a monopoly ads business, and a $54.2 billion / year YouTube business (set to eclipse Disney this year as the biggest media company in the world).

Google initially rode an undercurrent of demands for privacy in the wake of several high-profile privacy events: the Cambridge Analytica scandal, GDPR coming into effect, and Apple's App Tracking Transparency framework release and its subsequent privacy marketing push. Google may have initially thought they could accomplish dual goals of gaining consumer goodwill just like Apple by eliminating "tracking" and enhancing their business lines — similar to what Apple did with ATT. 

However, unlike Apple, they attempted to do all this just as they stared down the barrel of multiple antitrust trials for their ad tech and search businesses. Apple gleefully decimated advertising on iOS by virtually eliminating its advertising identifier, IDFA, and pushed billions away from advertising competitors into its own advertising business. Apple did this without a shred of US government scrutiny while Google stewed on the sidelines, unable to do the same thing with cookies to enrich itself.

While regulators in the United States have not taken issue with Apple's disingenuous privacy push, the Europeans have surprisingly made a stand.

France hit Apple with a 150 million euro penalty for self-preferencing via ATT with its required user choice prompts that use more gentle language in its owned and operated apps ("personalized advertising") as opposed to the harsh language required in non-apple apps ("Tracking") in the popup for non-Apple apps on iOS. Germany is currently investigating the same unfair practices of ATT as well.

Witnessing the heavily privacy-conscious Europeans see right through the user choice schemes of a giant technology company may have influenced Google's decision to abandon their own planned cookie prompt dialog in Chrome.

Most people may have also forgotten that we have also been waiting to see what Google would do with its mobile operating system, Android. Implementing a similar ATT-style prompt restricting access to its advertising ID, GAID now seems completely off the table. 

Even Apple has decided not to rigidly enforce its own ATT rules, while massive companies appear to be directly violating them. The cold calculus of using privacy as cover may no longer be a viable strategy for any giant tech company.

The looming remedies ordered for Google's monopoly verdicts could include spinning off components of their company, which undoubtedly influenced their cookie decision.

The court could order a significant restructuring of the company, which would require spinning off its Chrome browser, Android operating system, AdX ad exchange, GAM publisher ad server, and who knows what else. By punting any major decision until after the courts decide its fate, Google can develop the best strategy to maximize shareholder value for itself and its spun-out entities once it has all the information.

And so concludes the Google cookie saga. I've worked at three companies, had two kids, lived through a pandemic, launched and sold an ad tech newsletter, and developed gray hairs since Google first announced its intention to deprecate cookies — only to end up right back where we started.

I genuinely hope this is the last article I will ever have to write about the Privacy Sandbox and Google's decision to deprecate third-party cookies. I am truly irritated when I think about all the time we have wasted on the subject when we could have been innovating on other fronts — and I know I'm not alone.

Maybe after remedies are issued by the courts and implemented, we won't have to worry about a single company upending the entire advertising industry again.