THE B2B STACK | 16 May — LinkedIn becomes the B2B identity layer + Google fixes B2B lead bidding + Meta says targeting is solved + Marketers leash the agents
The week the platforms stopped pretending B2B was just a corner of consumer
B2B has been negated for years when it comes to advertising technology, but the tides are changing and there are signs of increasing maturity around B2B and adtech
LinkedIn are making off platform moves galore and they just became the universal B2B identity layer that every major walled garden has agreed to wire into. Google previewed its first Smart Bidding strategy designed specifically for long B2B sales cycles. Meta used its Performance Marketing Summit to concede that targeting is no longer the hard part. And the Digiday Programmatic Marketing Summit revealed what brands and agencies are actually doing with agentic buying behind closed doors, which is rather less dramatic than the keynote-stage version.
If you’ve spent the last decade running B2B media plans where the choice was always “LinkedIn’s audience or the open web’s reach”, this is the week the false choice started to disappear.
What you’ll learn in this issue:
Why the Amazon and LinkedIn CTV deal is bigger than another DSP partnership announcement, and what it means for B2B media plans that have run on LinkedIn alone
Why Google’s quiet pre-GML drop of journey-aware bidding matters more for B2B lead gen than the headline Performance Max updates
What Meta’s “targeting is solved” position at its Performance Marketing Summit tells us about the next eighteen months of platform competition
How real brands and agencies are running agentic ad buying right now, and the guardrails worth stealing
A separate read on what The Trade Desk’s Q1 earnings reveal beyond the share price
Four stories that actually move the road in front of you.
1. Amazon hardwires LinkedIn into its DSP. The B2B identity layer is now official.
The story. On 7 May Amazon Ads and LinkedIn announced that advertisers can target LinkedIn’s professional audience signals inside Amazon DSP for connected TV campaigns. Job title, industry, seniority, company data, all available to buy as deal-based CTV through Microsoft Monetize, LinkedIn’s parent SSP. The integration sits alongside Amazon DSP’s existing supply partnerships with Netflix, Roku, Disney, Spotify and SiriusXM, which together cover ninety per cent of US households via authenticated graph technology. Amazon led its third annual upfront at the Beacon Theatre four days later. The Trade Desk became LinkedIn’s first programmatic CTV partner earlier this year. Amazon DSP is the second. No other major DSP currently offers LinkedIn professional data at CTV scale.
Why it matters. For fifteen years the B2B media plan has been a forced choice. You bought LinkedIn for the audience and paid the LinkedIn tax. Or you bought the open web for the reach and lost most of the B2B identity at the door.
The Trade Desk and Amazon DSP moves end that choice for CTV specifically. LinkedIn is no longer a destination buy. It is a portable audience signal that the largest non-Google DSPs have now agreed to honour as a first-class targeting layer.
That changes the maths on a B2B media plan immediately. LinkedIn’s reach inside its own feed has always been capped by feed inventory and dwell. Their attempts to extend off platform with a checkbox for their audience network (LAN) was always clunky and they seem to be ceding that they need partnerships to scale this, letting them focus on being the B2B identity layer
LinkedIn audiences on a ninety-per-cent-household CTV graph are not capped by the same thing. They are capped by your willingness to pay LinkedIn the data premium across more inventory.
The B2B-specific read. Two structural shifts are happening at once. First, LinkedIn is becoming the de facto identity layer of B2B, the way Google’s logged-in graph became the identity layer of consumer addressable advertising a decade ago. Second, Microsoft Monetize is doing the unglamorous SSP work that makes this portable. The combination has been telegraphed in plain sight since Microsoft shut down Microsoft Invest in 2025 and named Amazon DSP as the preferred transition partner. This week was the productisation moment.
For B2B buyers, the strategic question is not whether to use this. It’s how to layer your own signal on top. LinkedIn data is professional identity. It’s not buying intent. Stacking high-quality intent and engagement signal underneath the LinkedIn audience, through partners that operate at the raw signal layer rather than at processed scores, is where the next round of competitive advantage lives. The teams that win 2026 won’t be the ones with the most LinkedIn budget. They’ll be the ones that understand LinkedIn audiences are now a commodity layer everyone has access to, and that the differentiation has moved one floor up the stack.
2. Google’s quiet B2B move: journey-aware bidding lands in beta.
The story. On 7 May, two weeks ahead of Google Marketing Live, Google’s Ads team dropped three bidding and budgeting changes via the Ads and Commerce Blog.
The headline-grabber is journey-aware bidding, a beta Smart Bidding strategy for Target CPA Search lead gen campaigns that values the entire customer path, not just last-click events.
Smart Bidding Exploration is also being expanded to Shopping and Performance Max. And demand-led budget pacing is coming to Search and Shopping. The lead gen change is the first Smart Bidding strategy Google has designed explicitly with multi-touch B2B sales cycles in mind and is long overdue
Why it matters. Anyone running B2B paid search has known the dirty secret of Smart Bidding for years. The Target CPA models optimise for the last clicked event before a conversion, which in B2B almost always under-credits the long-tail mid-funnel research queries that actually shape the deal. The result was Google preferentially spending against branded search and bottom-of-funnel intent, which is also the cheapest stuff to win organically. The forecasts looked great. The incremental pipeline rarely did.
Journey-aware bidding rebalances that. By giving algorithmic weight to earlier touchpoints in the path, the model can theoretically value a “what is X” query the same way a sharp human buyer would, as part of an opportunity that closes nine months later.
The B2B-specific read. Two cautions. First, this is still Google’s model fed by Google’s view of the customer path and will almost definitely bias to Google touch points, as they always do. The deeper your CRM feedback through enhanced conversions, the more accurate the journey reconstruction. Teams that haven’t invested in offline conversion uploads or enhanced conversions API integration won’t get the benefit. Second, this is a Search product, not a programmatic one. It tightens the loop on Google’s lead gen channels. It doesn’t help anywhere your audience is researching outside Google’s properties, which in B2B is most of the buying journey.
The bigger story is what this signals. Google is now openly designing for B2B as a distinct customer with distinct measurement needs. That hasn’t been true for most of the last decade, where B2B advertisers have made do with consumer-grade Smart Bidding logic and worked around it with manual rules. Expect more from this thread at GML on Wednesday.
3. Meta admits targeting is solved. Creative is the new bottleneck.
The story. At Meta’s Performance Marketing Summit in San Jose this week, the dominant message across nearly every customer session was a quiet retreat from the targeting-as-product narrative.
Andromeda, Meta’s retrieval and sequence-learning system, is being framed as a settled capability. The new infrastructure investment is in creative.
AI-generated assets, creator-content remixing, dynamic format adaptation. The tension Meta couldn’t resolve on stage: AI-generated creative is being defaulted for smaller advertisers, while every session positioned authenticity and creator trust as the competitive moat. The same AI face available to your scaling DTC brand is available to every dropshipper in the same vertical.
AI is enabling platforms to target significantly more granular and small micro segments because of the power of its data mining - but creative limitations, which have always been a major issue even before AI, are all the more obvious now. Performance is being held back by a lack of granular messaging in the creative layer
Why it matters. For most of the platform era, the targeting layer was the moat. Match the user, deliver the impression, take the margin.
Meta is now publicly conceding that the targeting layer is commoditising, partly through their own Andromeda investment, partly through the broader agentic shift across the industry. What’s left to compete on is the creative input quality. That’s a much harder, much less defensible position to take if you’re Meta.
The B2B-specific read. It would be easy to read this as a consumer-only story. It isn’t. The same dynamic is arriving in B2B but on a delay.
LinkedIn’s Thought Leader Ads and the broader creator-as-credibility integration are the B2B equivalent of “real face, real voice, scaled by AI”.
this is why we are seeing B2B influence expand; trust, micro message segmentation across hundreds of very specific experts, and LLM farming
B2B platforms are heading to the same place.
The targeting is going to converge across the major channels (see story 1), and the differentiation will come from the creative input.
For B2B teams that have outsourced creative to brand agencies on long cycles, that is a real structural problem. The teams already building creative-as-infrastructure, point of view, repeatable formats, native testing capacity inside the platforms, are pulling ahead. The teams still running quarterly creative refresh cycles are going to find their CPMs walking up.
The deeper point is that “AI creative” isn’t the answer for B2B either. The same dropshipper-face problem applies to generic AI-generated B2B stock.
Sleek dashboards, faceless office workers, an “innovate confidently” pull-quote. Buyers tune it out faster than they tune out a banner. The teams winning will be the ones with proprietary point of view, named human authors, and a creative engine built to spin variants of that, not generate yet more average from a prompt.
4. The agentic buying mood shift: from FOMO to guardrails.
The story. The Digiday Programmatic Marketing Summit took place 6 to 8 May in Palm Springs, with the editorial recaps landing through this week.
The keynote-stage version had everyone deploying agents end-to-end. The behind-closed-door reality, captured in the recap pieces, is more grounded.
OMD sees a path to letting AI agents make optimisation decisions in H2 2026, but with mandatory human sign-off. Bayer is running agentic optimisation under hard spend caps and human approval on any recommendation. Kelly Scott Madison Media’s SVP of analytics framed the real concern plainly: an agent that hallucinates a CPM could blow a quarter’s budget in a weekend, and the agency, not the agent, carries the bag.
Separately, brand and agency execs aired live frustrations with Performance Max, The Trade Desk, and Meta’s Advantage+, with the line “Google doesn’t care that it’s terrible” doing the rounds in summit coverage.
Why it matters. The narrative that agentic buying is “fully here” was a January-through-March story. With the usual hyperbolic adtech swagger and the reality being way behind the narrative
The May reality is that it is here in narrow, supervised pockets. Reporting and QA workflows, audience research, pacing checks, supply-side QA.
Agents are doing real work in those tasks. End-to-end campaign autonomy is still being held back by exactly the right questions: hallucination risk, contractual liability, measurement validity.
The B2B-specific read. For B2B teams running through agency partners, this is the moment to write the guardrails into your contracts, not after a budget blow-up. The shortlist worth stealing from the brand-side rules in play: hard spend caps per agent, human sign-off on any recommendation above a threshold, no agent autonomy on creative approval, and an audit trail on every change to bidding or pacing logic. None of this is hard. It’s a one-page addendum to your IO or MSA. Most B2B teams haven’t asked for it yet because most B2B agency relationships are still slightly behind the agentic curve.
Worth noting: the IAB Tech Lab’s Agentic RTB Framework and the Ad Context Protocol that came out in late 2025 are the rails this whole conversation is running on. Both are worth understanding before your agency tells you “we’ve turned the agent on”.
Worth a separate read: TTD’s Q1, and what Samantha Jacobson’s departure to OpenAI actually signals
The Trade Desk’s Q1 numbers hit on 7 May. $689 million revenue, twelve per cent year on year, beating consensus. EBITDA margin compressed from thirty-four per cent to thirty per cent. Gross margin at seventy-four per cent, an eight-quarter low. The stock dropped 1.75 per cent on the day, capping a decline of more than seventy-seven per cent from its 52-week high.
The number that mattered less than the headline was the timing of CSO Samantha Jacobson’s exit to OpenAI as VP of Partnerships for Monetisation, confirmed to Adweek hours before the print. She stays on TTD’s board. That follows the earlier exits of CFO Alexander Kayyal and CMO Ian Colley.
The Trade Desk’s response was the right one. Jeff Green spoke warmly about Jacobson personally, talked up the “biggest buyer’s market in the history of advertising” thesis, and pointed to JBP momentum and ninety-five-per-cent-plus client retention. The fundamentals beneath the share price are still very real.
But the talent flow is the story. Senior adtech leadership moving into OpenAI’s monetisation team in the same week OpenAI opened its self-serve Ads Manager with CPC bidding to all US advertisers is the most concrete signal yet that the AI platforms are building real ad businesses with real adtech operators inside them.
The Open Internet’s leading independent platform has just lost its strategy lead to the platform most likely to fragment the open internet’s intent-formation layer over the next three years. That is not a coincidence, and it isn’t going to be a one-off. Watch the next senior move.
Forwarded this email? Subscribe below
The B2B Stack lands every Friday. No recaps, no five-things-to-watch slideware, no vendor press releases dressed up as insight. Just the implication layer of what actually moved in B2B that week, written by someone who has been operating in B2B programmatic for fifteen years and has watched most of these cycles play out once before.
If you’re rethinking how signals turn into pipeline, that’s the problem we built FunnelFuel to solve. → Talk to us.
Sources
• Campaign US — Amazon Ads and LinkedIn partner to expand B2B reach for 2026 upfront, 5 May 2026
• MediaPost — Amazon, LinkedIn Team For B2B Inventory Ahead Of Upfront, 8 May 2026
• ppc.land — Amazon DSP integrates LinkedIn targeting for B2B streaming TV advertising, 8 May 2026
• Google Ads & Commerce Blog — Bidding and budgeting announcements, 7 May 2026
• ppc.land — Google targets hidden conversions with new bidding and budgeting tools, 8 May 2026
• GroAS — Today In Google Ads: May 14, 2026, Google Marketing Live 2026 Is Tomorrow, 14 May 2026
• Billo — Meta Performance Marketing Summit 2026: Targeting is solved. Meta is coming for creative next., 14 May 2026
• Digiday — Digiday Programmatic Marketing Summit May Recap: How marketers are navigating agentic ad buying, 15 May 2026
• Digiday — Marketers put up guardrails as AI agents reshape programmatic buying, 14 May 2026
• Digiday — The Trade Desk’s $689M revenue beat is blunted by the departure of CSO Jacobson to OpenAI, 8 May 2026
• The Motley Fool — The Trade Desk (TTD) Q1 2026 Earnings Transcript, 7 May 2026
• OpenAI — New ways to buy ChatGPT ads, 5 May 2026
• Deadline — Upfronts 2026 Report Card: Disney & Netflix Sizzle, NBCUni Fizzles, 14 May 2026