Google's Liz Reid on visibility in the AI search era

Google’s head of Search says personalization in AI search can lift niche and local sources, not just big publishers. Speaking on the AI Inside podcast on 26 June 2026, Liz Reid, VP and head of Google Search, argued that when Search understands what a user cares about, it can surface the specialist creator or the nearby shop that would otherwise be invisible. She also conceded that data on how AI answers drive clicks is still thin, because AI Mode and AI Overviews are both young.

What happened

The interview, hosted by Jason Howell and Jeff Jarvis, covered the friction between Google’s AI search experiences and the sites that supply the underlying content. Reid pushed back on the idea that personalization only helps large, established publishers. Her argument was that knowing a user’s interests, such as a preference for eco-friendly brands or the fact that they have a young child, lets Search connect that person to the niche reviewer, the local business, or the specialist that a generic ranked list would bury.

She was candid about measurement. Publishers have complained that Google provides little click data for AI Mode and AI Overviews in Search Console, and Reid framed this as an evolving area rather than a solved one, encouraging sites to build their own frameworks for measuring success against their own goals.

“AI mode is one year, AI overviews is two years. So we’re still learning.”

Liz Reid, Google, on the AI Inside podcast

She was also frank about the limits of Google’s own hold on users. If Search fails to connect people with the information they want, she said, they will simply go directly to the source, a reminder that visibility and usefulness, not just ranking mechanics, decide where attention lands.

Why it matters

Two signals stand out for anyone responsible for search visibility. First, personalization is being positioned as an opportunity for specific, distinctive sources rather than a threat to them. That favours businesses that are genuinely relevant to a particular person in a particular place, which is exactly the position a well-managed local presence occupies. Second, the measurement gap is real and acknowledged at the top of Google Search. Brands cannot assume AI surfaces will hand them clean click attribution any time soon.

The tension the interview kept returning to, less referral traffic and less data even as AI answers grow, is the same one enterprise teams are living with now. Reid’s framing does not resolve it, but it does point to where brands still have room to act: relevance, distinctiveness, and being the obviously useful answer for a specific user.

What this means for multi-location brands

For a brand with hundreds or thousands of locations, personalization surfacing niche and local sources is good news only if each location is actually distinctive and accurate in the data. A generic, half-complete location profile gives the system nothing specific to match a person to. The work is to make every location genuinely answerable.

Give each location real, locally specific content and accurate data so personalization has something to surface, and manage that consistently with a local business listing source of truth and AI-ready profiles through Places AI. Because AI surfaces will not report clean clicks, build your own measurement: use location and search insights to track calls, direction requests, and visits per location rather than waiting on full attribution from Search Console. Then treat this as the same discipline across engines, whether you are working to rank in AI search results or approaching generative engine optimization, the input is the same accurate, distinctive local presence.

The bottom line

Google’s head of Search is telling brands two things at once: AI search can surface specific, local, distinctive sources, and it will not give you full data on how it happens. For multi-location operators the response is not to chase the black box but to own the inputs, make every location genuinely useful and accurate, and measure the outcomes that matter on your own terms.