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

Three out of four advertisers are already seeing real performance gains from AI. But almost all of those gains are happening inside Google and Meta, and nowhere else.

A new study from Taboola surveyed 200 senior performance marketers managing monthly ad budgets between US$500,000 and US$4.9 million. The findings are clear: AI is working. It's just working exclusively inside the two platforms that already dominate the internet.

The result is a structural trap. Advertisers know it's a problem. Most want out. But the exit is blocked.

AI Tools Are Supercharging the Platforms That Already Win

Google's Performance Max is used at full scale by 91% of the marketers surveyed. Meta's Advantage+ is used by 88%. Both platforms have been aggressively deploying AI to automate targeting, creative, and budget decisions.

Marketing's Real Scale Challenge: Leadership, Not Tools

Organizations that scale creative work effectively distinguish themselves through clear decision-making systems, not tools. The real bottleneck is leadership, not technology.

Mission MediaMission Media

Meta's AI-driven ad system is growing at 24.1% annually, compared to Google's 12%. Meta is forecast to surpass Google in global ad revenue for the first time in 2026, reaching US$243 billion against Google's US$239.5 billion. The AI tools are delivering results inside these closed ecosystems. So advertisers keep feeding the machine.

But they're increasingly uncomfortable doing so. Brand and agency executives have publicly vented frustration with Performance Max and Advantage+ for hiding settings, auto-enabling targeting features without notice, and offering what one executive called "flying blind" transparency. That same frustration surfaced in a Digiday investigation into Performance Max and Advantage+ opacity, where one executive's verdict was blunt: "Google doesn't care that it's terrible." Despite that frustration, the money keeps flowing.

The Problem Isn't Willingness. It's Infrastructure.

Here's what makes the Taboola research striking: it isn't that advertisers want to stay locked inside Google and Meta. It's that they can't get out.

82% of the marketers surveyed say they're ready to adopt AI-driven campaign systems that work beyond search and social. 80% say they'd immediately increase spending on the broader internet if comparable AI tools existed there. 86% would shift up to 25% of their entire performance budget there under those conditions.

The blockage is operational, not strategic. 74% cite vendor complexity as the main barrier. 71% point to fragmented measurement systems. 54% struggle to integrate new AI tools into existing workflows without disrupting what already works. The AI tooling that makes Google and Meta so effective simply doesn't exist elsewhere at equivalent quality yet.

The APAC Dimension

For marketing leaders across Asia, the situation has an added layer of complexity.

52% of B2B Tech Leaders Now Prefer AI Search Over Google

52% of B2B leaders now prefer AI search over Google. A structural shift reshaping how Asian marketing teams reach buyers.

Mission MediaMission Media

88% of mobile ad budgets in APAC flow to Google and Meta, even though those platforms capture only about 30% of mobile users' attention. In mobile-first growth markets like India, Indonesia, and the Philippines, the gap between where audiences spend their time and where budgets go is widening.

McKinsey finds that well-deployed agentic AI in marketing can deliver 5% to 10% incremental revenue growth and 15% to 20% efficiency gains. But those figures remain out of reach for advertisers without the infrastructure to deploy AI beyond platforms that come pre-built for it.

What Could Change This

Taboola launched a beta AI system called Realize+ in April 2026, designed to bring automated budget reallocation and creative optimization to open web inventory. It's the kind of infrastructure parity that could shift behavior if it delivers.

Measurement is the other variable. Research shows first-click performance on the open web can generate three to four times more sales than last-click attribution alone. Most advertisers use last-click measurement, which systematically understates the value of any channel that isn't the final step before a sale.

By 2027, Google, Meta, and Amazon are projected to control 83% of global digital ad revenue outside China. Reversing that requires the open web to close the AI tooling gap. For APAC marketing leaders, the question isn't whether to diversify. It's whether the infrastructure will arrive in time.