The trend: Quarterly earnings from Amazon, Google, Meta, Booking Holdings, and a cluster of ad tech firms paint a consistent picture—AI now touches nearly every part of the marketing stack, from search and discovery to campaign production. 

Within advertising and commerce specifically, it still counts for a small slice of revenues. AI infrastructure businesses like Google Cloud are a different story, already a major and fast-growing chunk of the top line, because they provide the pipes that run AI processes.

That gap between AI adoption and revenues plays out differently depending on where AI sits in the funnel. Two areas show it most clearly: how people find products and how they buy them.

Zooming in: Consumers are finding products through chatbots and AI-powered search well before that traffic converts into sales at scale, and the shift is already rewriting how brands compete for visibility.

• Google’s Universal Commerce Protocol (UCP) links retailers like Target and Steve Madden to AI shopping experiences. Separately, Google says 500,000 advertisers now use its AI Max Search campaign tool.
• Booking Holdings says Google’s AI Mode and AI Overviews are cutting into traditional search referrals, prompting the company to test paid ads inside ChatGPT.
• Reddit faces a similar dynamic: Its content feeds AI answers without always driving visits back to the platform. CEO Steve Huffman counters that direct app users are worth far more than search referrals.

Recommendations for marketers: Visibility in AI-driven discovery is outpacing revenue from it, which means the brands that can build the right measurement and data habits might be better positioned once the channel matures and adoption catches up.

• Clean up product data first. Catalog listings are complete, but scraped ones often miss key attributes. Google's UCP runs on the same principle: Structured, accurate feeds beat whatever an AI agent can piece together from a webpage.
• Track AI referrals as their own line item, not a subset of organic or paid search. Volume is still small, but Amazon and Shopify data shows it converts differently.

Marketers who begin measuring now, while adoption is still ahead of revenues, will have clean baseline numbers on conversion rate, order value, and referral volume—before paid placements arrive and make it harder to tell whether an “AI referral” purchase came from strong product data or a paid spot.

This content is part of EMARKETER’s subscription Briefings, where we pair daily updates with data and analysis from forecasts and research reports. Our Briefings prepare you to start your day informed, to provide critical insights in an important meeting, and to understand the context of what’s happening in your industry. Non-clients can clickhereto get a demo of our full platform and coverage.