adMarketplace Data Shows AI Search Ads Converting 156% Better Year-on-Year
The Development
adMarketplace released results from the first full year of its AI Discovery alpha, launched in October 2025. Beta advertisers saw a 156% increase in conversion rates, and more than 30% of AI-chat queries were classed as commercially viable for advertising. Median query length was 8 words on browsers, 5 on comparison engines, and 3 on Buy Now Pay Later shopping apps. Electronics led on browsers, apparel on BNPL apps, and furniture on comparison engines. In the browser channel, up to 65% of queries ran seven words or longer, and 65% of those longer queries were transactional. adMarketplace framed the data against Google's search antitrust remedies, targeting a syndication market it sizes at roughly $100 billion.
Our Take
Treat the 156% with discipline: it is vendor-reported, from a beta cohort, against a low early baseline. The durable signal is structural. Query behaviour differs by environment, so an 8-word browser prompt and a 3-word BNPL search are different buying moments needing different creative, product feeds, and bids. That ends the assumption that AI search is one channel. It is a portfolio of intent surfaces, and the first budgets will be won by teams that segment accordingly. The antitrust angle matters too. If Google's remedies loosen syndication access, independent marketplaces become a real second source of high-intent inventory, which changes negotiating leverage on search spend heading into 2027.
What Changed
Advertisers can now buy against conversational intent across independent AI surfaces, not just keywords inside a single platform. Long prompts carry budgets, preferences, and requirements, so ad matching shifts from keyword proximity to context-rich intent, with measurable conversion data by surface type.
Marketing Impact
Paid search teams must extend keyword logic to prompt-level intent mapping and build surface-specific creative and product selection. Ecommerce and performance marketers gain a new inventory source with category skews they can plan against. Analytics teams need to separate AI-surface conversion reporting from legacy search to avoid blended-average blindness.
Competitive Implication
Performance teams that segment by AI surface gain a cost and relevance edge while competitors run one generic approach. Independent search publishers and syndication marketplaces gain bargaining power if antitrust remedies hold. Brands wholly dependent on a single search platform's ad stack become vulnerable as intent and inventory fragment.
Strategic Outlook
Expect conversational ad inventory to scale through 2027 as more publishers launch AI search experiences and Google's remedies shape syndication access. Independent benchmarks will follow, likely tempering the 156% headline, but surface-level intent differences will hold. Budget share for AI search will grow from test allocations into planned line items by mid-2027.
The Exploit
Action Item
Heads of paid search should carve out a Q1 2027 test budget, split across browser, comparison-engine, and BNPL AI surfaces, with separate creative and product sets per surface, and report conversion by surface before any scaling decision.