Display Is B2B's Most Undervalued Channel + ChatGPT Might Just Fix That. Plus: Trade Desk Courts OpenAI, LinkedIn Tells You to Cut the Bullspend, and the Google Ruling Could Drop Any Day
The B2B Stack's Weekly Intelligence Briefing
The Trade Desk’s worst month just got complicated.
Down 33% for the year. Three holdcos out the door in thirty days. The CEO spending $148 million of his own money defending the narrative. And then, right at the bottom of the selloff, a leak drops that TTD is in early talks with OpenAI to power advertising inside ChatGPT.
The stock bounced 18% in a single session. The founders $148m stock buy back suddenly looks a heck of a trade.
That’s the TTD story in microcosm right now: structurally under pressure, narratively fascinating, and impossible to write off. The OpenAI talks don’t fix the holdco problem. They don’t resolve the Publicis audit. They don’t rebuild the agency relationships. But they do point at something important — the next major ad inventory surface is being built right now, and whoever gets a seat at that table early shapes the decade.
Meanwhile LinkedIn showed up to NewFronts this week with genuine swagger, a campaign called “Cut the Bullspend,” and a pitch that’s as much pressure on every other B2B media platform as it is a product announcement. The Google AdX remedy ruling — overdue since Q4 2025 — could now land any day. And the IAB Tech Lab quietly unified all its agentic work under a single initiative, signalling that the standards war for autonomous programmatic is entering a new phase.
This week was the TTD-OpenAI story. But around it, a set of signals that, when read together, tell you exactly where the B2B advertising stack is heading.
Here’s what they mean.
📩 STACK SIGNALS lands in your inbox every Friday. The week’s B2B intelligence briefing, distilled.
THE PATTERN
There’s a single thread running through every major story this week.
Trust.
The Publicis/TTD fight is fundamentally about who trusts whom in the programmatic supply chain — and who has the leverage to demand accountability. The LinkedIn “bullspend” campaign is a direct attack on the industry’s long-running habit of substituting metrics for outcomes, and the trust gap between marketing teams and their CFOs that creates. The OpenAI/TTD talks are about whether a completely new ad surface can be built on top of existing infrastructure — which only works if advertisers trust it enough to actually spend there. There is also the consumer trust element - does the average ChatGPT user trust the platform when the answer is surrounded by ads? The Google ruling is about whether structural trust can be legislated into a market that’s been structurally compromised for a decade.
When the whole industry is arguing about trust simultaneously, it’s not a coincidence. It’s a reckoning.
STORY BREAKDOWN
STORY 1: The Trade Desk, Publicis, and the War Nobody Is Being Honest About - which has rumbled on from last week as expected
What happened: In an email sent to select clients and obtained by Adweek, Publicis explained that an independent audit by FirmDecisions found TTD had improperly applied its DSP fee to other charges, billed clients for tools they were automatically opted into without evidence of authorisation, and failed to provide the auditor with information necessary to validate that media and data costs were invoiced at cost and without mark-up. Ouch. Adweek
This followed Dentsu and WPP, who had stepped back from TTD’s OpenPath supply initiative weeks earlier. Trade Desk stock has retreated 33% in 2026. Campaign Live
Three of the Big Five holdcos. One month. All citing transparency. Publicis accounts for more than 10% of TTD’s gross billings, with two ad holding companies each accounting for over 10% of 2025 gross billings — combined, around 30%. Investing.com
Why it matters - my read:
Take the audit findings at face value. Fee complexity, auto-enrolment, audit access refused. Real problems. TTD’s fee structure is transparent in principle but confusing in practice. It has been slow, stubborn even, in response to legitimate concerns about the Kokai interface. Adweek Those are genuine operational failures worth fixing.
But they don’t explain the pattern.
Three holdcos, same month, same language. That’s not three independent audits arriving at the same conclusion. That’s a coordinated moment. And the commercial logic underneath it is visible if you look for it.
Industry observers pointed to potential overlap between OpenPath and services traditionally handled by agency trading desks, including supply-path optimisation and inventory curation. This is getting to the heart of the matter. e4m OpenPath wasn’t just a transparency initiative — it was TTD moving into the supply chain and threatening the margin line that agency trading desks charge for SPO and curation. When WPP cited “overlap with agency services,” they weren’t confused about what OpenPath does. They were telling TTD to get out of their lane.
The audit gave them a legitimising narrative for what is fundamentally a power move. You can’t tell clients “we’re stepping back because it competes with our trading desk margin.” But you can absolutely tell them “an independent auditor found multiple violations.”
The irony — which nobody is saying loudly enough — is that the loudest voices on transparency are the same organisations facing their own transparency crisis. WPP is currently subject to a $100 million whistleblower lawsuit alleging nearly $1 billion annually in undisclosed principal trading income. Publicis’s own holdco model runs on disclosed margins and undisclosed recovery. To survive, agencies have to generate profit from hidden fees and undisclosed arrangements. Marketers, in turn, reward this by continuing to hire the agencies that play the game most aggressively. Adweek
That is not a small irony. It’s the defining contradiction of the current holdco moment.
My view:
Stop reading this as a billing dispute. Read it as a territory war over who controls the supply chain in a post-OpenRTB world. The holdcos want the SPO and curation advantage back. TTD tried to disintermediate them. Neither side is being fully honest about why.
The B2B buyers in the middle should be asking a single question: if my DSP and my agency are publicly fighting over who controls the path to inventory — what does that mean for the quality, transparency, and true cost of the media I’m buying? The answer is more than most teams currently account for.
None of this is about the foundations of TradeDesk as a world class DSP. It boils down to lanes and margin.
STORY 2: TTD and OpenAI in Talks — The Most Important Sentence in Adtech This Month. Display Advertising Has Been B2B's Most Undervalued Channel for a Decade. ChatGPT Just Changed That.
What happened: OpenAI has been running early advertising trials since February, exploring partnerships with ad technology providers to support its effort to diversify revenue beyond subscriptions and enterprise services. ChatGPT reaches an estimated 910 million users globally. Early indications suggest OpenAI may use CPM pricing — a model that aligns with TTD’s strengths in inventory valuation and automated bidding. TheDesk.net
OpenAI is in early talks with The Trade Desk about a partnership that would help the AI firm sell ads within ChatGPT — TTD’s infrastructure could be used to quickly scale up ChatGPT’s ad business, rather than OpenAI building walled gardens that rival Meta and Google. eMarketer
The stock jumped 18.4% on the news. The market read it as a TTD recovery story. That’s the wrong read.
The real story — and it’s a MUCH bigger one:
Let’s talk about display.
For the better part of a decade, display advertising has been the unloved stepchild of the B2B marketing budget. CMOs direct spend toward events, LinkedIn, paid search, and content syndication. Display gets the scraps — typically a small retargeting budget, managed by someone junior, optimised to CTR metrics that everyone quietly knows don’t mean anything, and measured by last-click attribution that gives it no credit for the pipeline it actually influenced.
The reasons given are always the same. Display doesn’t perform in B2B. The targeting isn’t precise enough. The audiences are too broad. The CPMs don’t justify the reach when you’re selling to a committee of seven people at 500 accounts.
There’s truth in parts of that. Generic banner display on the open web, bought without signal and measured badly, is a waste of money in B2B. Agreed.
But here’s what that argument has always missed: the problem was never the channel. It was the signal quality, the supply quality, and the measurement framework sitting underneath it. The channel itself — reaching a relevant professional at the right moment in their decision-making process, with a CPM-based buy that doesn’t require a click to work — is one of the most powerful tools in B2B marketing. The world’s most sophisticated B2B vendors figured this out years ago. Their in-house programmatic teams run display at serious scale, with serious signal, and generate serious pipeline from it. The channel works. It just requires expertise most B2B marketing teams don’t have in-house — and most agencies haven’t been incentivised to build.
So they defaulted to LinkedIn. And events. And search. And called it a strategy.
Now here’s where it gets genuinely interesting.
If ChatGPT becomes a display channel, bought on CPM, powered by TTD’s infrastructure, served into an environment where 910 million users are actively researching and asking explicit questions, the entire calculus flips.
Think about what that surface actually is from a B2B perspective. A buyer researching CRM vendors, cloud infrastructure, logistics software, or any other enterprise category isn’t scrolling passively. They’re asking specific, high-intent questions and reading the answers with full attention. That’s not a banner impression on a news site. That’s something closer to the intent moment of paid search — but on a CPM model, at scale, in an environment that the open web cannot replicate.
The brand safety argument that has always haunted display disappears — you know exactly where the ad appears and what content surrounds it. The targeting precision argument weakens — query-level context tells you more about buyer intent than any third-party segment. The measurement argument shifts — you’re buying presence at the highest-attention moment in the research process, which is exactly where B2B influence actually happens.
If this plays out, display advertising doesn’t just get rehabilitated in the B2B CMO’s budget. It potentially becomes the most strategically important channel in the stack — the place where you reach buyers at the exact moment they’re forming category opinions, with verified context, at a price point that scales.
The B2B-specific angle nobody is saying:
Millennials and Gen Z now make up roughly 70% of B2B decision-makers. These are the buyers who run their research through AI assistants rather than search engines. They don’t fill in content syndication forms. They don’t attend three-day trade shows. They don’t respond to SDR cold sequences. They ask ChatGPT which vendors to shortlist and let the answer do the work.
If that’s where the research happens — and the evidence increasingly says it is — then the brand that isn’t visible in that environment is not on the shortlist. Full stop. No amount of LinkedIn spend or event presence compensates for not existing in the research moment.
Display, done right, in the right environment, with the right signal — has always been the answer to that problem. Most B2B CMOs just never built the capability to run it properly.
ChatGPT as a display surface might be the forcing function that finally changes that.
The honest caveat:
A partnership with OpenAI might not last long. OpenAI is reportedly also developing its own adtech capabilities, and in a worst-case scenario could become a future competitor to TTD rather than a partner. AOL
So TTD’s window is real but potentially short. More importantly, the pricing model, the brand safety standards, the targeting parameters, and the measurement framework for ChatGPT advertising are all undefined. This is genuinely early. No B2B CMO should be reallocating Q2 budget toward it yet.
But as a directional signal about where high-intent B2B display inventory is going? This is one of the most important stories of the year. The channel that B2B CMOs wrote off is the one that may define their 2027 strategy — and the teams building display capability now, with proper signal and proper measurement, will have a structural head start when that inventory opens up.
The question to ask your team this week isn’t “should we try ChatGPT ads?”
It’s “do we actually have a display capability worth scaling if the inventory becomes available?”
Most B2B marketing teams, honestly, don’t.
STORY 3: LinkedIn Tells the Industry to Cut the Bullspend — and It’s Not Wrong
What happened: At their second-ever NewFront, LinkedIn showed up with genuine ambition — emphasising creator partnerships, live-event integrations, CTV distribution, and measurable outcomes. Their own ad campaign, “Cut the Bullspend,” launched across New York the same week, running on billboards, branded cars, and digital surfaces outside competitor NewFront venues. Marketing Brew
The campaign challenges the industry habit of optimising B2B video for metrics like clicks and impressions that look impressive on a dashboard but fail to deliver business results. BrandLink campaigns see an average 130% higher video completion rate compared to standard in-feed ads, and members exposed to BrandLink are up to 18% more likely to become a lead. Adweek
Why it matters:
“Cut the Bullspend” is genuinely sharp positioning. It names something every B2B marketer thinks privately and nobody says publicly — that a significant portion of their budget is producing measurable activity and unmeasurable impact, and that the metrics being reported to the CFO are increasingly implausible as proxies for pipeline.
Millennials and Gen Z now make up roughly 70% of B2B decision-makers, and what moves them isn’t noise or hype but credible expertise and confidence that a choice will drive real outcomes. Adweek LinkedIn is building a pitch that speaks directly to that generation — they don’t download whitepapers, they watch video from voices they trust, in environments where professional context does real work.
My read;
LinkedIn is not your friend. It’s a platform with a commercial interest in you spending more money inside its walls. “Cut the Bullspend” is good positioning, but read the logic carefully — the proposed remedy to bullspend is more LinkedIn spend, not less spend overall. Its already an over-saturated channel which has great data but low intent, and functionally under-reaches many personas.
That said, the underlying argument is correct. B2B marketing has had a measurement problem for years, and AI is making it worse by creating more surfaces, more touchpoints, and more attribution complexity simultaneously. The CFO pressure is real. The move from optics to outcomes isn’t optional in 2026.
The B2B teams that handle this well won’t just shift budget into LinkedIn. They’ll build measurement frameworks that can attribute pipeline influence across the full stack — including the dark funnel activity that never shows up in Campaign Manager. LinkedIn is one piece of that. A well-run piece. But it’s not the whole answer, and the “Cut the Bullspend” campaign knows that.
STORY 4: The Google AdX Ruling Is Coming — Probably This Week or Next
What happened: Judge Brinkema is expected to issue her ruling on Google’s ad tech monopoly remedies before the end of Q1 2026 — meaning it could land any day. The DOJ wants Google to sell its ad exchange AdX and potentially its publisher ad server DFP. The remedies trial concluded in November 2025. Linos NEWS
The market has been waiting five months. The delay itself has been a signal — the longer it takes, the more likely a complex, nuanced ruling that neither side fully wins.
Why it matters:
Whatever Brinkema decides, the structural question it answers — or fails to answer — is the same one the holdco/TTD fight is about: who controls the supply chain?
If AdX is divested, the entire open-web publisher supply stack is in play. SSPs, curation platforms, and deal marketplaces get a once-in-a-generation opportunity to capture inventory that has been locked inside the Google pipe for fifteen years. That’s not theoretical. Publishers who have never interacted with programmatic through anything other than AdSense would suddenly be accessible via competitive auction.
If the ruling is behavioural only — which is increasingly the likely scenario based on the judge’s scepticism during closing arguments — the structural monopoly remains intact, but with mandated interoperability. Prebid integration, UPR reform, data sharing. Meaningful on paper. Google will test the wording of every clause. Progress will be slower than the industry hopes.
My read:
Build your supply strategy like the ruling changes nothing. Because even in the best-case divestiture scenario, the legal process runs for years — Google appeals immediately, the market fragments gradually, publisher migration takes longer than anyone models. The curation layer is the right answer regardless. Deal-based buying with curated, B2B-quality supply paths — where you own the signal, control brand safety, and see the full supply chain — is not a hedge against the Google ruling. It’s the correct strategy in every scenario. The ruling just makes it more urgent.
The real question for B2B buyers isn’t “what does Brinkema decide?” It’s “am I building supply relationships that don’t depend on the answer?”
STORY 5: IAB Tech Lab Unifies Its Agentic Work — The Standards Battle Has a New Name
What happened: In March 2026, IAB Tech Lab unified all of its agentic advertising work under a single name: AAMP — Agentic Advertising Management Protocols. This brings together the Agentic RTB Framework (ARTF), Agentic Audiences, Agentic Mobile, and the Agent Registry, launched March 1st, under one umbrella. ARTF promises to reduce bid request-response latency by up to 80%, cutting round-trip times from 600-800ms to approximately 100ms. IAB UK
Rather than blowing up the stack, IAB Tech Lab is betting that agentic systems can run on top of familiar foundations like OpenRTB, AdCOM, and others, while layering in newer protocols such as MCP and Agent2Agent. AdTechRadar
Why it matters:
The IAB unifying under AAMP is a power move as much as a standards exercise. AdCP — the rival framework backed by Scope3, PubMatic, and others — has been gaining momentum as an agent-native alternative built from scratch. The IAB’s message with AAMP is direct: you don’t need a new stack. The existing stack can be agentified. Build on what’s already there.
Both arguments have merit. AdCP is cleaner and purpose-built for agent-native workflows. AAMP carries the weight of existing adoption and the trust of a standards body that’s been governing programmatic since the RTB era.
My read:
This standards fight matters less for what it decides technically and more for what it signals commercially. Whichever framework wins, the outcome is the same for B2B advertisers: AI agents are going to be making media buying decisions, negotiating deal terms, and optimising campaigns without human sign-off. That’s not 2029. Over half of brands are using or planning to use agents in 2026. AdTechRadar
The question is not whether your agency or DSP is building agent capability. They all are. The question is whether your first-party data and signal infrastructure is clean enough for an AI agent to act on it correctly — because a badly structured audience or a polluted intent signal doesn’t just underperform when a human is watching it. It compounds at machine speed and at scale. Garbage in, garbage out, at 100ms per cycle.
🧠 ZOOM OUT: What this week actually tells us
Shift 1: The supply chain is the new battleground — and everyone knows it. TTD vs the holdcos. Google’s remedy. IAB AAMP vs AdCP. LinkedIn building its own walled inventory. Every major story this week is fundamentally about who controls the path from audience to impression. The buyers — brands and B2B marketing teams — are largely spectators in a fight that directly determines the cost and quality of their media.
Shift 2: Intent is moving off the open web — and nobody has fully priced that in. The OpenAI/TTD story isn’t a DSP recovery narrative. It’s a signal that the dominant research interface for B2B buyers is shifting. If intent forms in AI assistants rather than on web pages, the entire signal layer that B2B programmatic runs on is pointing at the wrong place. The industry is building faster, more agentic infrastructure on top of a signal model that may be structurally eroding.
Shift 3: Measurement credibility is now a competitive advantage. LinkedIn’s “Cut the Bullspend” resonated this week because it named something true. CFO scrutiny on marketing spend is at a generational high. The teams that can demonstrate causal pipeline impact — not dashboard metrics — will get budget. The ones that can’t are already having difficult conversations. Incrementality measurement isn’t a nice-to-have. It’s table stakes by Q3.
⚙️ WHAT SMART B2B TEAMS DO MONDAY
Audit your agency agreement. If you’re running programmatic through a holdco, ask explicit questions. What is the gross cost vs net cost of every line? Is principal trading or proprietary inventory involved? The Publicis/TTD fight just gave you the justification to ask questions you should have been asking for years. Use it.
Map your intent signal sources. If your ABM motion depends on third-party intent from web-based behavioural signals, model what happens to that data if 20% of your ICP’s research moves into ChatGPT or Perplexity within 18 months. Because it’s already happening for some buyer segments.
Build measurement infrastructure now. Not a dashboard. Not a report. A methodology — ideally always-on incrementality — that you can take to your CFO and defend against scrutiny. The “bullspend” problem is real, but the solution isn’t switching platforms. It’s being able to prove, causally, what’s working.
Run LinkedIn with intent, not dependency. The platform is exceptional right now — better targeting, better formats, better measurement than it’s been. But the walls are getting higher. Run it hard while ensuring you’re building signal outside it simultaneously.
Don’t wait for the Google ruling. Start building curation-first supply relationships this quarter. Whatever Brinkema decides, it will take years to implement and Google will fight every word of it. Your supply strategy can’t wait for a court case that’s been running since 2023.
💡 Next week: The Dark Funnel Playbook — a full activation guide for building pipeline from the 80% of your ICP that will never fill in a form, click an ad, or identify themselves anywhere in your current stack. The accounts that are already deciding before you can see them.