AI-driven performance gains are now measurable across most advertiser portfolios, but the automation infrastructure delivering those results remains concentrated inside Google and Meta platforms, according to Taboola research published May 15. The study found that 76% of advertisers report AI-driven performance improvements, yet adoption of intelligent campaign systems at scale drops sharply outside the search-social duopoly.

The Taboola report titled “The agentic advantage in performance marketing: Securing incremental growth beyond search and social” surveyed performance marketers on AI tool adoption and channel allocation patterns. While 82% indicated readiness to adopt goal-based AI campaign systems that drive outcomes beyond traditional channels, actual deployment figures reveal a structural bottleneck: Google Performance Max is used at scale by 91% of respondents and Meta Advantage+ by 88%, but open web AI campaign solutions operate at scale in only 36% of advertiser environments.

Budget allocation follows the same pattern. The research shows 74% of advertisers allocate at least a quarter of performance spend to paid search, with 67% doing the same for paid social, while open web channels receive comparatively moderate investment despite stated demand for diversification.

Marketing dashboard showing AI campaign performance metrics across search, social, and open web channels

Operational Barriers Block Open Web Expansion

The constraint is not performance skepticism but workflow friction, according to the study. Integration difficulty topped the barrier list at 54%, followed by vendor complexity at 74% when evaluating open web platforms and fragmented measurement systems at 71%. Brand safety concerns registered at 54% and limited internal resources at 42%.

Senior decision-makers and high-spending advertisers showed the strongest urgency, with 75% describing incremental performance discovery beyond search and social as very or extremely important. The demand signal becomes concrete in investment intent: 86% of respondents said they would allocate up to a quarter of performance marketing budget to open web channels if AI-powered automation matched the capabilities currently available only on walled-garden platforms.

TikTok’s Smart+ product represents a middle ground in the adoption curve. While widely tested, it has not yet achieved the scale deployment rates of Google and Meta offerings, the research indicates. This suggests that even well-resourced platforms outside the duopoly face structural challenges in matching the workflow integration and attribution clarity marketers expect from AI campaign systems.

Budget Reallocation Hinges on Infrastructure Parity

When asked to model budget allocation under a scenario where open web environments offered AI-driven optimization comparable to search and social platforms, respondents projected an average 24% allocation to open web performance channels. At least 39% indicated they would increase that share to 26% or higher, representing material reallocation from current distribution patterns.

The study framed the gap as structural imbalance rather than channel preference. According to Taboola, AI is delivering measurable gains but primarily within closed ecosystems where automation is most mature. The next phase of performance marketing will likely depend on whether open web environments can achieve measurement parity, attribution clarity, and workflow integration at the level currently offered by Google and Meta.

A broader Southeast Asia pattern emerged in parallel McKinsey research. While AI adoption is now widespread across the region, most organizations struggle to translate usage into measurable business value, the McKinsey report found. Although companies have begun using AI across at least one business function, a much smaller share have successfully scaled it into core operations. Even in markets with higher adoption rates such as Singapore and Indonesia, many organizations remain in pilot or early-stage deployment.

The Taboola findings align with earlier research the company released showing 86% of performance advertisers would shift a quarter of budget to the open web if agentic AI solutions delivered parity with walled-garden capabilities. The May 15 report provides deployment and barrier data that explain why that shift has not yet occurred at scale.

For enterprises operating digital marketing for ecommerce programs across APAC, the concentration risk is operational as well as strategic. When campaign automation, attribution modeling, and creative optimization infrastructure exist primarily inside two platform ecosystems, performance marketing leadership teams face limited leverage in negotiations and constrained options for incremental growth testing.

What This Means for — CMOs

Marketing leaders evaluating performance marketing agency partners in 2026 now confront a duopoly problem disguised as an AI opportunity. The automation gains are real—three-quarters of your advertiser peers report measurable improvement—but those gains currently require operating inside Google and Meta’s closed ecosystems. When you brief an agency on performance marketing expansion, ask specifically how they propose to access AI-driven optimization outside Performance Max and Advantage+, and what measurement infrastructure they use to maintain attribution clarity across open web placements.

The integration and vendor complexity barriers reported in the Taboola study signal a capability gap most agencies have not yet closed. If your agency proposes open web performance spend without demonstrating workflow integration with your existing attribution stack, you’re likely funding a pilot that will remain subscale. The 86% of advertisers willing to reallocate budget exists as latent demand precisely because the operational infrastructure lags platform readiness.

Budget planning cycles for H2 2026 and 2027 should account for this structural reality. Diversification beyond the search-social duopoly requires investment in measurement infrastructure and workflow integration before media spend scales. Agencies that lead with platform access rather than infrastructure readiness are selling you exposure, not performance.