Many changes that were already underway became permanent parts of the market. Search traffic did not return to previous levels, AI rapidly moved from experimentation into everyday use, and publishers’ advertising businesses began operating under tighter constraints. A single shift did not drive uncertainty, but the combined impact of multiple structural developments did.
Programmatic advertising continued to grow, particularly in premium environments and CTV, but growth did not necessarily translate into easier operations for publishers. Buyer requirements tightened, curation increased, and partially immature CTV structures added operational complexity. At the same time, AI Overviews, answer engines and chat-based interfaces began diverting audiences away from publisher content, reshaping traffic dynamics and, for some publishers, weakening the value of traditional traffic.
AI quickly became part of advertising optimisation, analytics and daily workflows. Focus shifted toward traffic quality, measurement reliability, and how advertising value is actually distributed between publishers and other market participants. At the same time, regulation, legal processes and rising transparency demands from buyers began to influence publishers’ decision-making more directly.
In this review, we examine the most essential AdTech themes of 2025 from the perspective of publishers’ advertising businesses.
Zero-click and the new reality of discoverability
The zero-click phenomenon became part of everyday reality as AI Overviews, answer engines and chat-based interfaces increasingly satisfied user intent without page visits. For publishers, this decouples visibility from traffic, directly weakening the logic of traditional display and programmatic advertising.
Search traffic not only declined, but also became harder to predict. This increased pressure on inventory management, pricing and measurement, particularly for publishers whose revenues relied on volume. At the same time, it became clearer that not all traffic is commercially equal, as attention, engagement and environmental quality increasingly determined outcomes.
Zero-click also highlighted the question of value distribution. Publishers’ content fuels search platforms and AI-generated answers, but the benefit does not return as traffic or advertising revenue. This accelerated discussions around AI visibility, licensing and different compensation models as part of publishers’ long-term business strategies.
AI in publishers’ day-to-day operations: operational support
AI became an established part of publishers’ daily operations. The conversation shifted from experimentation to practical use, with AI applied especially in reporting, optimisation, content analysis, and as operational support for ad sales and Ad Ops teams. It helped process large volumes of data and detect anomalies faster than manual processes.
However, the benefits remained largely operational. AI accelerated analysis and decision-making, but did not solve publishers’ core structural challenges, such as declining traffic, fluctuating inventory value, fragmented measurement or dependence on external platforms. In practice, AI operated within boundaries set by publishers: it could optimise floor prices and suggest adjustments, but it did not determine which inventory is sold, how it is packaged, or under what terms buyers gain access.
Over the year, it became clear that AI’s value for publishers lies primarily in control and responsiveness. AI helped monitor market changes and prioritise actions, but strategic decisions around commercial models, partnerships and maintaining control remained firmly in human hands.
Programmatic advertising tightens
Programmatic advertising has evolved into a more demanding environment. Growth increasingly focused on premium inventory, private deals and CTV, while the volume-driven logic of the open market lost some of its relevance. For publishers, this meant a shift from optimising quantity toward tighter control over quality, governance and commercial terms.
At the same time, buyer expectations increased. Transparency, transparent pricing, cleaner bidstreams, and a better understanding of what makes inventory valuable and why became central requirements. This was reflected in stronger supply-level curation, wider adoption of SPO thinking, and growing interest in new standards such as the Deals API.
Publishers gradually realised that programmatic advertising works best when it is actively managed. Continuous monitoring, testing, and structural decisions became more critical as programmatic sales alone no longer performed as they once did. When implemented correctly, however, programmatic remained a core component of advertising businesses even in a tightening market.
Curation, deals and supply-path thinking
Curation increasingly moved away from individual PMP packages toward structural supply-path thinking.
Buyers began defining more precisely which inventory is allowed into auctions in the first place and at what stage buying decisions are constrained. In practice, this meant inventory was filtered and prioritised also before auctions rather than during them. This shift was reflected in curation moving closer to the SSP level, ad servers, and especially CTV environments, where deal-based trading remains central.
At the same time, the role of deals changed. Manual deal IDs and isolated packaging proved inefficient and difficult to manage for many, increasing pressure for automation and shared structures. IAB Tech Lab’s Deals API and growing interest in supply-path optimisation reflected buyers’ need for clearer, more predictable routes with fewer intermediaries. For publishers, this meant that managing the supply path was no longer just a technical issue, but directly tied to pricing, revenue distribution and trust with buyers.
In 2025, curation, deals, and supply-path thinking converged into a single framework in which control, clarity, and operational efficiency outweighed volume. Publisher success was no longer built on the broadest possible distribution, but on the ability to offer buyers a clean, understandable and commercially viable route to high-quality inventory.
Regulation, Google and transparency
Regulation and market power moved from background discussion into everyday publisher decision-making. Legal proceedings related to Google’s ad tech business in the U.S. and Europe, the implementation of the Digital Markets Act, and debates around the roles of browsers and operating systems created an environment in which stable rules could no longer be taken for granted.
The impact was not seen as a sudden disruption, but rather as a call for caution and reassessment. Publishers were forced to reconsider their dependence on individual platforms, the resilience of their measurement frameworks, and the extent of absolute control they could retain over their businesses. Uncertainty itself began influencing investments, partnerships and technical choices.
At the same time, transparency increasingly became a matter of trust. Both regulation and buyer expectations moved in the same direction, toward simpler structures, more transparent accountability and a better understanding of how value is created and who ultimately captures it. For publishers, 2025 made it clear that market power, regulation, and transparency are not separate topics but core elements of long-term sustainability in the advertising business.
Summary
In 2025, publishers’ operating environments did not become simpler, but their constraints became more visible. Traffic no longer behaved as before, programmatic sales required more active management, and AI settled into everyday use primarily as a support tool rather than a shortcut to growth. In practice, this meant more measurement, more testing and more deliberate choices in a context where automated assumptions no longer held. Where traffic comes from, how inventory is sold, and how much control publishers retain over supply paths began to impact results directly.
For many publishers, 2025 made it tangible how difficult it is to steer advertising businesses without clear visibility into structures, data and commercial terms. The challenge was not a single technology or model, but the ability to understand the whole picture and act consistently in an environment where predictability declines while expectations continue to rise.