Thirty-eight percent of consumers now use AI to research products during the festive shopping season, according to data from Integral Ad Science (IAS). That figure should reset how product managers at consumer brands think about discovery: AI assistants are no longer an edge case in the purchase funnel — they are a mainstream entry point.

The number comes from IAS, the ad verification and media measurement company, as reported by MediaBrief. It quantifies a shift that many commerce and growth teams have sensed anecdotally: shoppers preparing for festive-season purchases are delegating early-stage research — comparing options, surfacing recommendations, narrowing consideration sets — to AI tools rather than typing queries into a search bar and clicking through results pages.

For practicing PMs, the implication is concrete. If more than a third of your prospective buyers are forming their first impressions of your product through an AI-generated answer, then the content and structured data you publish — specs, pricing, availability, use cases — becomes machine-readable inventory for someone else's summarizer. Pages optimized for human skimming and SEO ranking may lose fidelity when an intermediary model compresses them into a two-sentence recommendation.

There are real failure modes here, and they cut both ways. AI tools can misstate prices, omit a key differentiator, or recommend a competitor in the slot where you used to rank second organically. Brands that treat AI-driven discovery as a channel to monitor — auditing what major assistants say about their category, correcting stale or wrong data at the source — will be better positioned than those that discover the drift after a holiday quarter's conversion numbers disappoint.

The measurement question is unresolved, which is the honest caveat. IAS's 38% figure is a research statistic, not a channel-attributed revenue number, and definitions matter: "using AI to research products" can span everything from asking a chatbot for gift ideas to reading an AI-generated review summary on a retail site. Product teams instrumenting for this shift should decide early what they count — AI-referred traffic where identifiable, AI-influenced surveys, or share of assistant answers mentioning their brand — because the metric choice will drive very different roadmap priorities.

What is not ambiguous is the direction. Festive-season behavior tends to foreshadow year-round habits; capabilities shoppers adopt under gift-buying time pressure tend to persist into routine purchasing. The PMs who spend this quarter mapping how AI assistants describe their category — and fixing the upstream data those answers draw on — will be doing the least glamorous, most compounding work of the AI-commerce transition.