MarTech Cube with Emma Lacey on the next phase of programmatic, focusing on AI, smarter optimisation and real business outcomes.
Since the genesis of real-time bidding nearly 20 years ago, the programmatic ecosystem has been developing at a breathless pace. What began as a solution to the problem of shifting publishers’ unsold inventory, rapidly became a complex marketplace where billions of ad impressions are auctioned in a matter of milliseconds. Now, the programmatic system for the open web is far too fragmented. The industry keeps adding layers, which means more intermediaries, more platforms and more complexity for buyers to deal with.
However, progress is being made as the emphasis shifts from merely selling volume of impressions to driving meaningful outcomes. Programmatic’s next phase is about maximising measurable business impact from media investment. This has been enabled by advances in exchange AI technology and sell-side intelligence that are making it possible to optimise campaigns at scale in real time. With data points covering everything from placement to performance available instantly, it’s now feasible to build a single optimisation loop that lets marketers see the meaningful, measurable impacts their campaigns have.Â
How the programmatic open web reached this turning point
For all its innovation, programmatic has been beset by problems over the years. Ad fraud is a persistent challenge, brand safety crises have repeatedly occurred, and fragmentation issues mean that a single impression’s journey from advertiser to consumer could pass through a dozen intermediaries. Â
The result? The ANA’s 2023 Programmatic Transparency Study found that just 36 cents of every dollar entering a DSP was actually reaching a consumer, with the rest swallowed up by transaction costs, invalid traffic, non-viewable impressions, and made-for-advertising (MFA) sites built to game the system rather than serve audiences.Â
While the situation has improved since then, more change still needs to happen. The ANA’s Programmatic Transparency Benchmark from Q2 2025 showed that some $26.8 billion in global media value is still lost each year to inefficiencies. Despite this, programmatic investment is projected to grow to nearly $800 billion by 2028.
How the open web has caught up with the walled gardens
For years, walled gardens had a clear advantage over the open web, as they brought media and data together in a single system. Creative execution was part of the same workflow, offering advertisers simplicity and measurable results in one place. The open web, by contrast, operated across disconnected platforms with fragmented workflows, creating complexity that made it much harder to show the same value.
But now the picture is somewhat different. According to eMarketer data, walled gardens lost share of programmatic ad spend for the first time in 2024, and this trend is likely to continue as the open web becomes more competitive on outcomes. This trend is supported by integrated platforms with sell-side intelligence, which are giving the open web the same simplicity that walled gardens built their dominance on, but with the added benefit of demonstrable results.Â
The importance of publisher-led programmatic intelligence
Traditionally, optimisation happened inside the DSP, after the bid request had already arrived. By that point, much of the value had already been lost. The explosion of bidstream volume, combined with signal loss between demand and supply platforms, made it clear that optimisation needed to move earlier in the process.
Sell-side intelligence involves applying optimisation before inventory enters the bidstream. Quality signals and creative performance indicators are assessed at the impression level, filtering out low-quality supply and enriching the value of high-quality supply. Buyers get better performance, and publishers get more value from their inventory.
To be effective, this approach must be applied to three key components.. A media quality engine applies real-time data to enrich each impression before the bid. A creative intelligence engine uses immersive formats, such as AR and 3D, alongside real-time performance signals, treating creative as a live optimisation input rather than a static asset. An AI decisioning layer then optimises at the impression level for outcomes across the full marketing funnel from brand affinity to conversions.
A better value exchange for the open internet
Bringing these three components together produces a unified platform that connects media quality, creative performance and business outcomes in one AI-driven optimisation loop. Advertisers can set real business goals, such as CPA (cost per acquisition) or CPL (cost per lead), and the system continuously adjusts supply decisions based on live performance data. This approach is capable of delivering a significant improvement in outcomes.
This is a meaningful change for publishers too. When supply is optimised for quality and performance, publishers benefit from stronger demand and better yields. Audiences also get a better deal, through improved ad experiences with more relevant creative formats.Â
A bright future for the programmatic ecosystem
The technology is now in place to simplify the open internet, so it can produce the outcomes it has long promised and, as the evidence shows, it is beginning to do exactly that.
For more expert articles and industry updates, follow Martech News
Emma Lacey, SVP Sales Demand EMEA, Onetag
Emma is a commercial leader with 20+ years’ experience in the advertising industry, specialising in programmatic and ad tech. She has built and led high-performing teams, secured strategic partnerships with global agencies and brands, and worked across the publisher, demand, and supply sides of the ecosystem, spanning contextual, video, and CTV.