This week I pulled Google Search Console data across our full portfolio of 24 real estate websites, layering in Google’s newer “Performance — Generative AI” reporting. The question: what kind of content is actually showing up when buyers use AI-powered search to find a home?

The data: 65,367 AI search impressions across 6,924 pages. Listing/property-search pages? Only about 10% of that visibility.

What outperformed them, per-page:

Comparison/decision content — 31.1 avg impressions

Costs/financial content — 30.8 avg impressions

Consumer questions/transaction content — 30.6 avg impressions

Substantive community/geography content — 25.1 avg impressions (Portfolio average: 9.35)

Inventory shows consumers what they can buy. Expertise helps them understand what, where, and how they should buy.

The takeaway isn’t “stop showing listings.” IDX inventory is still the foundation — buyers need homes to search. But the sites winning in AI search are the ones that also answer the questions buyers already have: What does it cost to live here? How does this compare to that other place? What should I know before I buy?

This is observational data, not a controlled experiment — association, not proof of causation. Next step: publish a batch of comparison, cost, and community pages and track generative visibility at 30/60/90/180 days.

Full analysis available on request.