86% of Advertisers Ready to Shift Budget—If Open Web AI Catches Up

Four in five large-scale advertisers say they would shift more budget to the open web tomorrow, if only the AI tools matched what Google and Meta already offer. That is the blunt finding from a new survey by Taboola of 200 senior performance marketers, each managing monthly ad budgets between US$500,000 and US$4.9 million.

The gap is simple to state. The big walled gardens (Google, Meta, Amazon) have spent years building AI systems that run campaigns almost automatically. Outside those platforms, on news sites, publisher pages, and the broader internet, nothing comparable exists. Advertisers know it. They want it. And a race is now underway to build it.

What the Numbers Actually Reveal

The survey makes the frustration quantifiable. 76% of marketers are already seeing better results from AI-driven advertising tools, but nearly all of those gains are happening inside Google and Meta, not across the broader internet. The open web gets 59% of people's online time but captures only 48% of advertising budgets. That 11-point gap represents real money on the table.

80% of Marketers Want to Leave Google and Meta—But Can't

80% of marketers want to escape Google and Meta, but infrastructure barriers—vendor complexity, fragmented measurement, and lack of comparable AI tools—keep budgets locked inside closed platforms.

Mission MediaMission Media

What makes this story more than a niche ad tech debate is the scale of the latent demand. 86% of those surveyed said they would shift up to 25% of their performance budget to the open web if the right automated tools existed. At these budget levels, that translates to hundreds of millions of dollars in potential reallocation.

Adam Singolda, CEO of Taboola, put the advertiser demand plainly: "Our research shows a clear demand from advertisers that want the same 'always-on,' AI-driven performance they see in walled gardens applied to the open web. They are looking for autonomous systems that learn continuously, pivot in real time, and turn every impression into a measurable outcome."

Why the Big Platforms Keep Winning

The platforms' dominance is not an accident. Meta's Advantage+ system, built on data from 3.2 billion daily active users, runs at a US$60 billion annual revenue run rate and helps drive Meta's projected overtaking of Google in global digital ad revenue in 2026. Together, Meta, Google, and Amazon are on track to command 62.3% of worldwide digital ad spending this year.

These platforms offer something open web vendors have struggled to replicate: AI that automatically adjusts creative, audience targeting, and budget allocation without requiring a human to log in and make changes. Advertisers set a goal. The system figures out how to hit it.

For a campaign manager spending US$1 million a month on ads, the appeal is obvious. The open web, for all its reach, still asks them to do more of that work manually.

Who Is Trying to Close the Gap

Several vendors launched agentic advertising products in early 2026, signaling the industry recognizes the urgency. Taboola rolled out Realize+, which includes a Decision Engine that shifts budget in real time and an Element Generator that automates ad creative, replicating the core logic of how Meta Advantage+ works. The company also integrated Realize+ with Anthropic's Claude, allowing agencies and advertisers to manage campaigns through a conversational AI interface, a distribution approach no major walled garden has matched.

Marketers Commit US$16 Billion to AI Search—But Doubt the ROI

Marketing leaders commit $16B to AI search in 2026, but 60% doubt the ROI. Budget outpaces organizational readiness and measurement infrastructure.

Mission MediaMission Media

PubMatic launched its AgenticOS in January 2026 with live campaigns already running on connected television. Yahoo DSP and Magnite have made similar announcements. IAB Tech Lab published a roadmap in January 2026 to extend the industry's programmatic buying standards for AI agents, partly to prevent a fragmented landscape where every vendor builds incompatible systems.

The Integration Problem Nobody Is Solving Fast Enough

Here is the catch that the survey surfaces clearly. The barrier to adoption is not skepticism about AI. It is operational complexity, and the problem is worse for the biggest advertisers.

Only 9% of marketers spending US$300,000 to US$499,000 a month flagged integration into existing workflows as their main obstacle. Among those spending US$1 million to US$4.9 million a month, that figure jumps to 74%.

Large organizations do not just plug in a new ad tool. They have internal approval chains, existing technology contracts, regional teams with different setups, and reporting requirements that do not disappear because a new platform arrives. Only 6.3% of companies have fully integrated AI agents into their marketing operations despite 90.3% claiming to use AI agents in some form.

For APAC marketers, this complexity multiplies. A regional campaign spanning India, Indonesia, South Korea, and Vietnam involves separate language requirements, different publisher relationships, and varied platform regulations. The promise of autonomous AI optimization is appealing, but the operational lift of making it work across fragmented regional stacks is the real challenge.

As Yahoo DSP's Adam Roodman put it: "Agentic AI shifts DSP competition away from the UI and towards performance." That reframing is accurate, and it is also a warning for vendors still competing on interface design rather than results.