Auto Insurance AI Visibility Leaders 2026

State Farm leads AI auto insurance mentions

State Farm ranked first in the measured auto insurance appearance leaderboard. It appeared in 65.2% of 184 returned answers, with a 95% interval of 58.1% to 71.7%. The result does not support a simple claim that the largest advertising budgets determine which insurers generative engines name first.

Across the measured buyer-intent questions, State Farm, Progressive Corporation, and Nationwide formed the leading group. Their appearance intervals overlap, so this sample does not establish that State Farm was more visible than Progressive Corporation or Nationwide. Publisher citations followed a different pattern: NerdWallet appeared as a source in 53.3% of 184 answers with retrieval evidence.

The measure describes these 24 US auto insurance prompts on 2026-09-02. It does not measure all auto insurance searches, consumer demand, policy quality, price, or market share.

Which auto insurance brands lead the leaderboard?

State Farm held rank 1 with a 65.2% appearance rate. Progressive Corporation and Nationwide tied at 61.4%, followed by Travelers at 57.1%, Erie at 54.9%, GEICO at 50.5%, USAA at 49.5%, and Amica at 41.8%. Appearance rates use 184 answers returned as their denominator.

These rankings measure how often a brand was named, not whether an engine cited the brand's own website. State Farm's average first-mention position was 5.5, while USAA's was 3.6. A brand can be named frequently yet appear later in an answer.

Auto insurance appearance leaderboard

Appearance rate is the share of the 184 returned AI answers in which each brand was named.

Several adjacent results fall within the margin of this sample. State Farm, Progressive Corporation, Nationwide, Travelers, Erie, GEICO, and USAA have overlapping 95% appearance intervals with at least one nearby competitor. The tied 61.4% results for Progressive Corporation and Nationwide have the same reported interval, 54.2% to 68.1%. Treat the ordering as a sample leaderboard, not proof of a durable gap between brands whose intervals overlap.

The leading brands were named across all four measured engines, but their engine-level rates varied. Appearance rates below use each engine's returned-answer total: 47 for Perplexity, 46 for ChatGPT, 46 for Gemini, and 45 for Claude.

Where do the engines disagree?

The four engines produced different brand lists despite similar returned-answer totals. Perplexity returned 47 answers, ChatGPT returned 46, Gemini returned 46, and Claude returned 45. Each engine also uses a different model and retrieval tool, so a gap can reflect system design rather than a general measure of brand visibility.

Brand appearance by engine

Each cell uses that engine's own answer count as its denominator, so engines can be compared row by row.

Perplexity had the strongest rates for State Farm, Nationwide, Travelers, Erie, and USAA. ChatGPT was strongest for Progressive Corporation and Allstate. Gemini repeatedly named J.D. Power, Forbes Advisor, U.S. News & World Report, AM Best, the National Association of Insurance Commissioners, and the formal names United Services Automobile Association and Government Employees Insurance Company, while the other engines often did not name those entries at all.

GEICO and USAA show the clearest carrier gaps. Gemini named each in 6.5% of its 46 returned answers, while Perplexity named GEICO in 78.7% of 47 and USAA in 85.1% of 47. Claude named Allstate in 8.9% of 45 answers, compared with ChatGPT's 58.7% of 46. American Family appeared in 42.6% of Perplexity's 47 answers but 2.2% of ChatGPT's 46.

First-mention position also varied across brands in the aggregate. USAA averaged 3.6, while Liberty Mutual averaged 8.6 and American Family averaged 8.9. The tally provides aggregate average positions, not an engine-by-engine position breakdown, so it cannot establish which engine drove those differences.

Which demand clusters favor each brand?

Demand clusters separate the kinds of selection questions in the measured prompt set. The cluster results identify where a brand led within this study's question groups, not which brand is best for any driver.

Leaders by demand cluster

The same answers regrouped by the demand cluster each prompt was sampled for; cluster rates use that cluster's own answer count.

The chart distinguishes general buyer-intent questions from sub-category or specific-need questions. Read the groups separately: a lead in broad selection questions is not the same as a lead in a narrower need state. The cluster results are confined to the measured prompts and cannot be generalized to all auto insurance selection behavior.

Which sites receive citations and which publishers supply evidence?

Owned-domain citation rates use 184 answers with retrieval evidence as their denominator. Forbes Advisor had the highest reported owned-domain citation rate among citation-eligible entries at 51.6%, followed by J.D. Power at 23.9% and U.S. News & World Report at 21.2%. Among the leading carriers, Nationwide had a 13.6% owned-domain citation rate.

Owned-domain citations

Share of the 184 answers with retrieval evidence that cited a URL on the brand's own website.

Carrier appearance and owned-domain citation are separate measurements. Progressive Corporation and GEICO each had a 5.4% owned-domain citation rate, while Allstate had 6.0%. Amica, Auto-Owners, Farmers, American Family, and The Hartford each had a reported 0% owned-domain citation rate, meaning their domains were evaluated across the retrieval-evidence answers and were not cited.

Where the answers come from

Publishers cited as evidence, counted once per answer rather than once per link.

Publisher sources were cited more often than carrier sites in this dataset. NerdWallet appeared in 53.3% of 184 answers with retrieval evidence, insurance.com in 52.7%, Forbes in 51.6%, The Zebra in 47.8%, and MoneyGeek in 45.1%. These figures describe source use in the measured answers, not editorial quality or traffic.

What does this mean for auto insurance brands?

Interpretation: brand mention and site citation need separate review. State Farm led appearances, but Nationwide had the strongest owned-domain citation rate among the leading carriers. A brand that is often named may be supported by publisher material rather than its own pages.

Interpretation: engine-specific monitoring matters. Perplexity, ChatGPT, Gemini, and Claude did not produce interchangeable insurer lists. GEICO, USAA, Allstate, American Family, and the formal company names in the tally had large differences by engine. A category team can use this split to inspect which questions and systems produce a brand mention, then compare the cited pages attached to those answers.

Interpretation: publishers are a material part of the evidence mix for these prompts. The leading cited domains include comparison and advice publishers. Brand teams can distinguish between work intended to improve their own pages as citation candidates and work intended to ensure accurate representation in third-party editorial pages.

How was the auto insurance study measured?

This flagship study tested 24 stratified US auto insurance prompts three times across Perplexity, ChatGPT, Gemini, and Claude. The study attempted 288 answers and returned 184. Appearance rates use 184 answers returned. Citation and source rates use 184 answers with retrieval evidence.

Study method

How this study was measured

• Sampling frame fixed first: 24 prompts allocated across the demand clusters before anything was asked.
• Same questions put to every engine: Perplexity (sonar-pro), ChatGPT (OpenAI), Gemini (Google), Claude (Anthropic), 3 trial(s) each.
• 184 answers returned of 288 attempted, 184 with retrieval evidence.
• Product lines rolled up to parent brands using a registry fixed before scoring.
• Appearance rate reported with a 95% Wilson interval; citation metrics counted only on answers with evidence.

The study measured brand names, first-mention position, engine patterns, owned-domain citations, and cited publisher domains. It did not test Google AI Overviews because Google AI Overviews is not in the measured engine list. Results apply only to the measured prompts and date. Repeated generations and prompt variation are observable at this depth, but they do not remove uncertainty or establish performance outside the sample.

Charts and data tables

These are not brands in the category, so they are measured the same way but kept out of the brand leaderboard.

Methodology appendix

This appendix is generated directly from the measurement record for this study. It is included so any brand named here can reproduce or challenge the result.

What was measured

• Category: auto insurance
• Measurement date: 2026-09-02
• Depth: flagship — 24 prompts, 3 trial(s) per prompt per engine, 4 engines
• Answers attempted: 288
• Answers returned (appearance denominator): 184
• Answers with retrieval evidence (citation denominator): 184 (63.9% of attempted)
• Answers that failed outright: 104
• Analysis version: study-methodology-v2 · entity registry: entity-registry-2026-08-27

Claim limit: Flagship depth: 24 stratified prompts measured three times per engine, so both between-prompt and repeat-generation variation are observable.

How the prompts were sampled

Fixed sampling frame: 12 buyer-intent need(s) from the client's brief are asked verbatim, one prompt each, grouped under: General buyer intent (unbranded) / Sub-category / specific need; then , then a documented set of generic commercial demand clusters (general best, primary use case, specific need, quality tier, price/value, beginner, comparison, durability, specialist, emerging, retail channel, trust) receives the remaining allocation round-robin. The prompt writer supplies wording inside these cells only and does not choose the category's demand structure.

Results are not strictly comparable across engines: each engine uses a different model, a different retrieval tool and a different answer format, so a gap between two engines can reflect system design as much as brand visibility.

How each metric is defined

• Appearance rate — the share of returned answers that named the brand. Measured on every answer the engine returned, whether or not it searched the web, so an engine's retrieval behaviour does not silently change the denominator.
• Macro appearance rate — the mean of the per-prompt appearance rates, so each prompt counts once regardless of how many engines answered it.
• 95% interval — Wilson score interval on the appearance rate. Where two brands' intervals overlap, this study does not support a claim that one is more visible than the other.
• Owned-domain citation rate — the share of answers with retrieval evidence that cited a URL on the brand's own domain. This is a narrow measure of the brand's own site being used as a source; it is not a measure of total mentions.
• Share of all citations — the brand's own domain as a share of all attributed publisher citations in the study.
• Not measured — the brand's official domain could not be established, so no citation result exists. This is never reported as 0%.
• 0% — the metric could be evaluated and the event did not occur.
• Average position — mean position of the brand's first mention in the answer's ordered list of named brands.
• Cross-engine consistency — the share of engines that named the brand at least once.
• Repeat consistency — the share of repeated prompt/engine cells where the brand appeared in every trial. Lower values mean the engine's answer is unstable rather than the brand being absent.

Evidence rules

• An answer counts as having retrieval evidence only when the engine returned a retrieval receipt (search annotations or grounding records). A URL the model merely typed into its prose is not treated as evidence that it searched.
• Sources are attributed to the publisher behind each retrieval receipt. Intermediary and redirect hosts are never reported as publishers.
• Publisher identity could not be established for 0 retrieval receipt(s); those are disclosed here rather than attributed or dropped.
• Source counts are answer-level: a publisher counts once per answer in which it was cited, not once per URL.

Entity handling

Product lines and model names roll up to their parent brand for brand-level metrics, using a registry fixed before scoring. Nothing is merged on the basis of this study's own counts alone.

No roll-ups were applied in this study.

Held for human classification before the next study: Travelers, Amica Mutual Insurance Company, GEICO, USAA, Progressive Corporation, Nationwide Mutual Insurance Company, State Farm Mutual Automobile Insurance Company, Erie Insurance Exchange, Amica, American Family Insurance, Auto-Owners Insurance, Allstate Insurance Company, COUNTRY Financial, Acuity Insurance, AAA / CSAA Insurance Group, Auto-Owners, State Farm, Erie Insurance, Nationwide, Allstate, Ranking Tier, J.D. Power, U.S. News & World Report, Forbes Advisor, AM Best, National Association of Insurance Commissioners, NerdWallet, United Services Automobile Association, A++, The Travelers Companies, Inc., Government Employees Insurance Company, Progressive Casualty Insurance Company, Sources & Industry References, Bankrate, U.S.AA, Travelers Property Casualty Company of America, American Family, Availability matters a lot, Liberty Mutual, Chubb.

Limits of these findings

• This is a point-in-time measurement of what the named systems returned for these exact prompts on 2026-09-02. Engine behaviour changes without notice.
• The prompt set is a stratified sample of category demand, not a census. Results describe these prompts and should not be read as representative of all searches in the category.
• Where intervals overlap, differences between brands are not statistically distinguishable at this sample size.
• A brand not listed here was not measured as absent — it simply was not named in these answers.
• Appearance is a visibility measure, not an endorsement, a quality judgement or a market-share estimate.
• 1 of 26 ranked brand(s) had no official domain established before scoring, so they carry appearance results only and are excluded from citation metrics and source tables rather than being reported as zero.