A Google AI Overview does two things with brands. It names them in the answer text, and it credits sources beneath it. We measure both separately, and they are frequently not the same companies.
Named more than cited
| Brand | Line | Named in answer | Cited as source |
|---|---|---|---|
| USAA | car insurance | 9.5% | 0.0% |
| Travelers | car insurance | 9.5% | 0.5% |
| Progressive | business insurance | 7.6% | 0.2% |
| USAA | home insurance | 7.2% | 0.0% |
| GEICO | business insurance | 15.6% | 10.1% |
Brand recognition is carrying these companies and nothing on their sites is reinforcing it. That position is only as durable as the model’s existing impression of them.
Cited more than named
| Brand | Line | Named in answer | Cited as source |
|---|---|---|---|
| Liberty Mutual | car insurance | 0.5% | 16.2% |
| Kin Insurance | home insurance | 0.0% | 11.5% |
| Progressive Commercial | business insurance | 7.6% | 17.8% |
| Liberty Mutual | home insurance | 0.4% | 10.4% |
| NEXT Insurance | business insurance | 8.3% | 17.5% |
These companies’ content is doing work they get no attribution for. The reader learns what they taught them and hears someone else’s name.
Neither position is visible in rank tracking. Both are measurable.
Method
2,900 insurance searches captured October 2026 through geo-targeted live SERP retrieval across 50 US states, 6 Canadian provinces and 109 metros. Brands and source types are attributed through a published entity taxonomy maintained by hand and republished each edition.
Full method: https://insuranceseosem.com/about/#methodology
The Index: https://insuranceseosem.com/index/