2026 · Q4 / Quarterly research edition
State of Local AI Search
Which local businesses appear in AI buyer searches? A cross-market view of recorded visibility, the services being measured and the evidence needed to assess content outcomes.
Observed through November 4, 2026 · Rolling 90-day data. This edition updates with observations; it is not a frozen quarterly study.
01 / The visibility gap
- Named at least once
- 29.1%
- Markets with checks
- 20 / 23
- Recorded firm checks
- 15,766
280 of 962 checked listings
14 configured services
Not unique answers or search volume
Unchecked listings are excluded from the named share. Firms appearing in multiple markets remain separate market listings. Coverage reflects our tracked sample, not every business in a city or the whole AI search market.
Metro snapshot
Six highest observed shares, with unequal sample sizes. Not a representative ranking of cities.
All 23 metros and sample sizes
02 / Services
Where businesses are being named
| Service | Checked | Named | Share |
|---|---|---|---|
| Home Inspection | 147 | 31 | 21.1% |
| Roofing | 95 | 26 | 27.4% |
| HVAC | 86 | 29 | 33.7% |
| Plumbing | 114 | 35 | 30.7% |
| Pest Control | 70 | 20 | 28.6% |
| Cosmetic Dentist | 63 | 17 | 27.0% |
| Estate Planning Attorney | 68 | 18 | 26.5% |
| Personal Injury Attorney | 81 | 34 | 42.0% |
| Water Damage Restoration | 62 | 9 | 14.5% |
| Foundation Repair | 68 | 17 | 25.0% |
| Real Estate Agents | 36 | 15 | 41.7% |
| Title Companies | 36 | 12 | 33.3% |
| Photographers | 36 | 17 | 47.2% |
| Independent Insurance Agents | 0 | 0 | Pending |
03 / Content outcomes
Is local content creating citation lift?
No completed, sufficiently measured local study is available for a lift claim yet. 0 active studies are currently public. Preparing studies remain private.
Frozen question/engine groups, at least 40 baseline checks, 3 untreated same-market/service controls, and 200 post-treatment checks per member. Net change = treated change − mean control change, in percentage points. These observational comparisons are not randomized causal proof, revenue attribution or a guarantee.
Study evidence and treatment pages04 / Sources & methodology
Follow the evidence, not assumptions
Source appearances, business mentions, crawler fetches, human visits and customer inquiries are different signals. No inquiry conversion or source-caused lift is inferred here. Historical SaaS results are not pooled with local studies.
CC BY 4.0 · Cite “PressGEO, State of Local AI Search,” the edition, observation date and this page. Engine adapters record returned sources with differing retrieval methods; those observations do not establish engine trust or endorsement.