PubMatic's Agentic AI Bet Optimizes Programmatic Delivery, Not Brand Trust
PubMatic just made a significant bet. The company announced an exclusive partnership with Untapped Growth to deploy AI agents that could change the programmatic supply chain -- no DSP required. The framing: AI agents can navigate or bypass the layers of intermediaries that make programmatic advertising expensive and opaque, creating a cleaner, more efficient path from buyer to inventory.
At roughly the same time, IAB Tech Lab signaled it's accelerating standardization of agentic AI workflows in advertising. Its integration with Kochava's StationOne was described as a shift "from AI hype to standardized, executable ad workflows." Two organizations, moving in parallel, with the same implicit diagnosis: complexity is programmatic advertising's core disease, and AI agents are the cure.
That diagnosis is coherent. It's also incomplete in a way that matters considerably to how brands should think about this.
What Programmatic Actually Industrialized
When programmatic advertising emerged around 2012-2014, the promise was precision. Instead of buying a television spot and hoping the right people saw it, you could target the exact person, at the exact moment, on the exact device, with the exact creative. This was a genuine technical achievement.
What it wasn't -- and never claimed to be -- was a solution to brand relevance.
Programmatic improved the efficiency of the pipe. It didn't improve what was flowing through it. Brands with weak value propositions got more efficient at distributing those weak propositions. Brands that had lost consumer trust could now lose it at scale, algorithmically, across thousands of publisher placements simultaneously. The complexity that IAB and PubMatic are now racing to solve with AI agents is partly structural (too many intermediaries extracting margin) and partly symptomatic: complexity accumulated because each layer in the stack claimed to add value that the underlying signal didn't actually have.
This is the same pattern Branding Strategy Insider identified this week in its analysis of Peloton, Kohl's, Target, General Mills, and Macy's -- all engaged in brand turnarounds, all making the same foundational error. The observation: "A product is not a brand. A product is the truth of a brand's promise." These companies are investing in product improvement as if better specs will automatically restore brand equity. They won't. The product is necessary but not sufficient. Brand equity lives somewhere upstream of product, in accumulated trust, in the coherence between what a company says and what it actually does.
AI agents in programmatic make the same category error. Better delivery infrastructure for a weak brand signal amplifies that weakness more efficiently.
The Supply Chain Optimization Trap
PubMatic's specific move -- deploying AI agents to enable programmatic buying that routes around DSPs -- is technically credible. Demand-side platforms add cost, latency, and opacity. If AI agents can replicate targeting and optimization while reducing these frictions, that represents a real efficiency gain for buyers.
But efficiency gains in ad tech have a reliable historical track record: they get competed away within 18 to 24 months, then get absorbed into the next layer of opacity that forms around them. This is what happened to real-time bidding. This is what happened to header bidding. Both were genuine technical improvements that temporarily reduced complexity before the market reconfigured and complexity returned.
The reason complexity keeps reasserting itself isn't that ad tech lacks engineering talent. It's that the market is structurally incentivized to add intermediary layers whenever the underlying brand signal is weak enough that every party can credibly claim they're adding value.
This is the kind of pattern our research tracks systematically -- technological improvements in advertising infrastructure that produce short-term efficiency gains but don't structurally alter the underlying brand trust dynamics that drive whether advertising actually works.
AI agents don't change the incentive structure. They change who executes the decisions -- and potentially who captures the efficiency savings -- but the gravitational pull toward added complexity returns whenever signal quality justifies it.
Why IAB Tech Lab's Standardization Push Is Different, and Still Insufficient
IAB Tech Lab's effort deserves to be distinguished from PubMatic's product bet. Standardization is a public good. When IAB creates interoperability standards for agentic AI workflows, it's not trying to capture margin -- it's trying to make the entire ecosystem function more predictably. That's legitimate and valuable.
But standardization solves coordination failure. It doesn't solve content quality.
A standardized, interoperable, AI-mediated programmatic ecosystem is better than a fragmented one. It is not better at ensuring that what moves through that ecosystem is worth receiving. A faster, more efficiently routed irrelevant ad is still an irrelevant ad, now arriving with lower latency and fewer margin leaks.
The McKinsey analysis we covered in the context of ITV's streaming collapse made a related point: as AI systems take on more autonomous action in media environments, the trust infrastructure required to operate in those environments fundamentally changes. The risk isn't primarily technical -- it's relational. Brands that haven't built genuine proximity with their audiences will find that AI-mediated buying delivers this deficit at scale and with increasing precision.
The Workforce Displacement Narrative Gets This Backwards
Nick Maggiulli at Of Dollars and Data pushed back this week on the "permanent underclass" narrative around AI. His argument: history shows no permanent underclass emerging from technological transitions. The workforce adapts, new categories emerge, and the feared scenario of permanent displacement from a single technology wave has never materialized at the scale predicted.
He's probably right about this. The argument maps interestingly onto brands navigating AI in advertising.
There won't be a permanent underclass of brands that can't access AI-powered programmatic tools. That technology will commoditize, as all ad tech does. The fear that smaller brands will be permanently disadvantaged by PubMatic-style AI optimization isn't the real risk -- large brands and small ones will all gain access to this infrastructure within a few years.
The real risk is that brands adapt toward AI-optimized delivery while their underlying brand equity continues to erode. You can learn to deploy every AI agent PubMatic ships and still be Macy's. You can run the most efficient programmatic operation in your category and still be Peloton -- a company where people can now identify the exact moment the product improvement strategy diverged from the brand recovery strategy.
Adapting to AI wins the tactical battle. It doesn't resolve the strategic one.
What the HBR Data on Employee Joy Reveals
This week Harvard Business Review published research showing that leaders systematically underestimate the value of employee joy. Organizations where employees experience genuine joy outperform on retention, customer satisfaction, and productivity -- but leadership consistently underinvests in the conditions that produce it because joy is difficult to measure and easy to deprioritize against more legible metrics.
This connects to the brand trust question in a way that isn't immediately obvious.
Brand authenticity -- the kind that survives AI-mediated advertising, that can't be optimized away because it's built into the operational fabric of how a company actually behaves -- is downstream of organizational authenticity. Brands where people are genuinely engaged tend to produce customer interactions that feel different. Not because of better creative, but because the humans involved believe in what they're doing, and that belief is legible even through automated delivery infrastructure.
Programmatic advertising optimized by AI agents can find the right audience at a lower CPM. It cannot manufacture the organizational energy that makes a message worth believing. If anything, the increased targeting efficiency of AI-mediated delivery makes the gap between authentic and inauthentic brands more apparent, not less. You're not hiding in broad demographic spray anymore. You're surfacing, precisely, to the exact people most likely to notice whether you mean it.
What the Smart Brands Are Actually Doing
We've argued before that brand strategy has shifted from promise to proof -- that brands don't build operational capabilities to support cultural positioning anymore, they earn cultural positioning by building operational capabilities. The infrastructure is the message.
AI agents in programmatic don't contradict this shift. They accelerate it.
If the AI agent can identify your most relevant audience with increasing precision, the question shifts entirely to what that audience finds when it arrives. A more targeted impression of a weaker brand is a more targeted disappointment. The brands that will compound through AI-mediated advertising environments are the ones building better destinations: products that genuinely solve problems, organizations where people experience something like joy in doing the work, brand trust that is operationally proven rather than aspirationally asserted.
PubMatic and IAB Tech Lab are building better roads. That's a legitimate infrastructure problem and they're right to work on it. The bet they're implicitly making is that complexity is the binding constraint on programmatic performance. It isn't. The binding constraint is signal quality. Reduce friction on a weak signal and you get a faster path to the same outcome.
The brands that will get meaningful ROI from agentic AI in advertising are the ones that have already done the harder work: clarifying what they actually stand for, building products that prove it, and creating the organizational conditions that make that proof legible to the people delivering it. AI agents help those brands find their audience. They don't create a message worth finding.
If you're evaluating how agentic AI infrastructure fits into your brand's advertising strategy, our analysis tools can help surface where the real leverage points are -- and where optimizing the delivery pipe is a distraction from fixing what flows through it.