Ask a media buyer why their ad won a given placement at a given price, and increasingly the answer is that nobody really knows.
IAB’s 2026 Digital Video Ad Spend report shows trust in inventory transparency is already fragile. Even the buying methods buyers consider most transparent, direct I/O, programmatic guaranteed, and self-serve platforms, are trusted by only 57% of buyers. Confidence drops below 50% for private marketplaces and to roughly a third for open exchange buying. Layer AI-driven decisioning onto an ecosystem that already struggles with trust, and the visibility gap risks widening for buyers and sellers alike.
Buyers aren’t the only ones losing visibility. Publishers face a version of the same problem from the other side of the auction: They can see their inventory bought, priced, and packaged by AI-driven systems, yet have no clear line of sight into why demand landed where it did.
Without visibility into which factors drive demand, a platform or publisher can’t confidently price or forecast fill rates. Publishers should be pushing for the same visibility buyers want, and standards like the MRC’s Auction Transparency Standards and IAB Tech Lab’s Programmatic Auction Definitions matter as much to the sell side as the buy side.
The world’s most powerful SEO platform, purpose-built for Enterprise.
From execution to intelligence
DSPs and SSPs enable the auction for both sides and feel this tension most sharply. Their edge was built on mechanics they owned: DSPs through bid optimization and targeting, SSPs through inventory access and monetization.
AI now automates much of that, making these systems more dynamic and harder for publishers and buyers alike to observe from the outside. When a model’s weights shift continuously, logic once at least partially inferable from outcomes becomes opaque.
Video and CTV feel this acutely, given how fragmented inventory is across platforms, devices, and buying methodologies. When a video ad loses an auction or a publisher’s inventory goes underbought, both sides need to know whether that reflects strategy or a technical issue on the platform side. Without visibility, everyone is left guessing.
This is pushing platforms toward new differentiators: proprietary decisioning and recommendation engines, predictive forecasting across fragmented supply, unique data signals, and the ability to show why an outcome happened rather than just report it.
Value in the supply chain
The pressure on intermediaries didn’t begin with agentic buying. A 2020 ISBA/PwC supply chain study found roughly 15% of ad spend untraceable to any line item, an unknown delta lost across too many hops. The finding intensified scrutiny around supply chain complexity and gave advertisers a concrete reason to ask whether every intermediary was creating measurable value.
The industry’s response has been years of supply chain rationalization and consolidation, along with demands for greater accountability.
One major adtech player wound down its advertising and data business in 2024, shuttering data and measurement units it had spent years and billions acquiring. It illustrated how difficult it can be for standalone data and measurement businesses to maintain differentiation as privacy changes, platform consolidation, and native platform capabilities reshape the market.
While these developments weren’t driven by AI, they foreshadow the same question AI is now accelerating across the video advertising ecosystem: Which functions create unique value, and which are increasingly becoming embedded in larger platforms and automated systems?
Ensuring durable advantage
Not every company sitting in the intermediary category faces the same future, and the industry’s response has split into several paths.
Owning more of the workflow
Functions that once required separate platforms are increasingly folded into larger, AI-enabled stacks. As AI takes on more inventory selection, optimization, and forecasting, intermediaries face growing pressure to prove they add value beyond simply sitting as another step between advertisers and media owners.
Owning unique signals
As AI makes optimization and decisioning more commoditized and widely accessible, durable advantage increasingly comes from owning something that can’t be easily replicated: proprietary data, unique inventory access, identity solutions, or specialized expertise.
Retail media networks are the clearest example of this dynamic in action: Their value comes from owning deterministic, closed-loop transaction data, exactly the kind of signal needed to differentiate on, and exactly the kind of asset that gets more valuable as commodity data gets squeezed out.
Owning trust and transparency
As AI-driven decisioning grows more complex, advertisers and media owners are placing more value on platforms that can explain their outcomes rather than simply report them. The ability to explain why an outcome happened is becoming a competitive differentiator, on par with efficiency and performance.
Bringing visibility back
The industry is building more transparency into programmatic transactions.
- IAB Tech Lab’s Open Measurement SDK gives buyers independent, standardized measurement of ad delivery and viewability, helping verify that impressions are valid before questions of pricing logic come into play.
- SupplyChain v1.1 aims to give buyers fuller visibility into the path an impression actually travels before it reaches a viewer.
- IAB’s General Terms & Conditions and the MRC’s Digital Advertising Auction Transparency Standards introduce clear reporting and accountability requirements to CTV transactions.
These efforts point to a broader shift: AI is increasing the value of assets and capabilities that intermediaries can uniquely own, from proprietary data and exclusive inventory to trusted decisioning infrastructure.
Transparency is becoming one of those assets. Platforms that can explain why an outcome occurred can give buyers and publishers something increasingly valuable: confidence in the decisions that drive their business.
The post Programmatic video needs more visibility into AI-driven decisions appeared first on MarTech.