Allstate: The Third-Highest Share of Model in the Study, and Rank Six Overall

1. The volume paradox

AIVI™ weights share of model at 35%, more than any other component. It is the measure of how much of the total organic mention volume in a category belongs to a brand, and Allstate is third on it in the entire 134-brand study cohort. On the heaviest single input to the composite, only State Farm and Progressive are ahead.

Allstate finishes sixth.

The two panels should step down together, and they nearly do — with one exception. Of the ten brands published, four move up between the two rankings and each of them moves up exactly one place. One moves down, and it moves down three. That asymmetry is the entire subject of this page: Allstate is the only brand in the top ten whose composite result is materially worse than its reach would predict.

This is the mirror image of the profile examined in the first teardown in this series. The Hartford holds the best position rate in the study and rank eight overall — precision without breadth. Allstate has the breadth. The question this page answers is what breadth is worth on its own, and the answer turns out to be less than most planning assumptions allow for.

2. What we measured, and what we did not

This section is deliberately placed before the analysis rather than after it.

• Public model outputs only. Every figure comes from organic brand mentions in responses generated by ChatGPT, Gemini, Perplexity, Copilot, and Claude. No advertising data, no analytics access, no internal reporting from any carrier.
• No client relationship. Brainpan.AI has no commercial relationship with Allstate. The company was not contacted, did not participate, and did not review this page. It appears here because it is the clearest case in the Q2 2026 top ten of reach failing to convert, not because of any dispute or engagement.
• Organic-only, first-mention attribution. Sponsored placements and paid surfaces are excluded. Where a response names several brands, attribution follows the first-mention rule described in the methodology.
• A sample, not a census. One quarter, 300 stratified prompts, 1,500 model responses. Prompt sets, model versions, and retrieval behavior all move between quarters. Treat quarter-over-quarter movement as directional until a second cycle confirms it.
• Open data, CC BY 4.0. The underlying benchmark is published under a Creative Commons Attribution 4.0 license. Reuse it, check it, disagree with it in public.

Two derived measures appear below. Visibility Points are the sum of position, recommendation, and citation scores; the line-of-business totals in section 6 are computed inside a single line and are not normalized 0–100 AIVI™ scores, so they should not be compared across lines of different sizes. Recommendation points weight a strong recommendation at 5 and a moderate one at 3; both weightings are set out in the versioned methodology.

3. Six point seven two

The cleanest way to see what the composite is actually pricing is to put Allstate next to the brand that finished one place above it. Nationwide appears in this series for a second time, and for the opposite reason: in the Hartford teardown it was the breadth case, and here it is the efficiency case.

Allstate leads the first five rows and loses the last five. It is named more often, holds more of the category’s mention volume, lands in the top three more often both in count and in rate, and places earlier on average when it appears. Then the table turns over, and the final margin is 1,439.12 Visibility Points against 1,445.84 — a gap of 6.72 points, or 0.47%.

Two line items account for the reversal. Nationwide records 18 more recommendation points and 36 more citations. Everything Allstate wins in the first half of the table is worth less than those two differences, which is a strong statement about what the underlying models are doing: being retrieved and placed well is table stakes, and the separation happens in whether the answer goes on to endorse the brand and whether it sources the claim.

4. The same 104 recommendations, twenty-four more mentions

Recommendation rate is the share of a brand’s mentions in which the model does more than name it — where the answer actively suggests it, ranks it, or steers the reader toward it. Allstate sits at 37.7% against a top-ten median of 40.2%. Nationwide sits at 41.3%. As percentages the gap looks modest. Converted into counts it disappears entirely, which is the more interesting result.

One hundred and four each. Allstate and Nationwide converted exactly the same number of mentions into active recommendations in Q2 2026. Allstate needed 276 mentions to produce that figure; Nationwide needed 252. Twenty-four mentions of additional reach bought nothing.

The 18-point difference in recommendation points comes from the composition of those 104. The benchmark separates strong recommendations from moderate ones and weights them 5 and 3 respectively. Nationwide’s 104 split 52 strong and 52 moderate, for 416 points. Allstate’s split 43 strong and 61 moderate, for 398. Half of Nationwide’s recommendations were unambiguous; 41% of Allstate’s were.

Across the whole top ten, Allstate returns 1.442 recommendation points per mention. That is eighth of ten, and last among the six brands with more than 250 mentions — the peer group that actually competes for the same answers. USAA, on six fewer mentions than Allstate, returns 1.930.

The counterfactual is the number worth carrying into a planning conversation. Hold Allstate’s mentions flat at 276, hold its 104 recommended mentions flat, and change only the strong share from 41% to Nationwide’s 50%. Recommendation points rise from 398 to 416, closing that line item outright — without a single additional mention, a point of share of model, or a new piece of content ranking anywhere.

5. Breadth without corroboration

The second line item is citations. Allstate records 83 across 276 mentions, a citation rate of 30.1% — ninth of ten in the published set, ahead of only State Farm, which offsets a 28.2% rate with 66% more mentions than anyone else.

The panels run in near-opposite directions, and that is not an accident of this quarter. The four smallest mention bases in the top ten all carry citation rates above 42%. Nationwide is the only brand in the study’s top ten holding a top-six mention volume and a citation rate above 40% at the same time. Reach and corroboration are being produced by different mechanisms, and a brand can be very good at one while sitting near the bottom of the other.

Part of Allstate’s rate is a structural ceiling that applies to every carrier. Sorting the full 134-brand cohort by entity type shows where citations actually land.

Five publisher domains hold 4.9% of the study’s organic mention volume and 11.6% of its citations. A publisher converts a mention into a citation 2.7 times as often as a carrier does. When an answer engine needs to source an insurance claim, it reaches past the carrier for a third party — which means a carrier’s citation ceiling is set partly by a structural preference it does not control.

That ceiling is the same for all 89 carriers, though, and it does not excuse the position. Allstate’s 0.301 citations per mention is barely above the 0.278 all-carrier average while carrying the third-largest mention base in the study. Within the top ten — the set of brands competing for the same answers — it is ninth.

6. Where the volume comes from

Aggregate share of model says how much of the category a brand holds. It does not say which part. Splitting the study by line of business is where Allstate’s profile becomes legible.

Allstate is second in Home on 429 Visibility Points and third in Renters on 200. It does not appear in the published top five for Auto, the line carrying the largest visibility totals of the five reported, where GEICO leads on 779 and Nationwide takes fifth on 293.

That is the finding. Allstate’s third-place share of model is assembled from the property lines, not from the line that generates the most answer volume in the category. Nationwide — the brand that beats it by 6.72 points — is the only brand in the study over-represented against NAIC 2025 market share in two separate lines, Auto at 7.1× and Commercial at 4.2×. Allstate appears in neither the over-represented nor the under-represented published lists.

Read together with sections 4 and 5, the constraint stops looking like a visibility problem in the ordinary sense. Allstate is present at scale. It is present in the wrong proportion across lines, and where it is present, the answer names it without committing to it and without sourcing it.

7. Strategic diagnostic priorities

Four moves follow from the data above. They are stated as diagnostic direction rather than as an execution plan — sequencing, prompt targets, and content specification are what an audit produces, and they depend on inputs no public dataset contains. The operating frameworks that carry that work — content standard, technical foundation, publishing protocol, and monthly cadence — are published in full at Brainpan.AI frameworks.

Stop treating mention volume as the constraint

Allstate already holds the third-largest mention base in a 134-brand study. A program that opens with increase AI visibility and measures itself on mentions is optimizing a variable that is already performing at the top of the category, and the twenty-four surplus mentions over Nationwide demonstrate what the marginal one is currently worth. The first question is not how to be named more often. It is what happens in the sentence after the name.

Diagnose the Auto line separately from everything else

A brand with the third-highest share of model in the study and no position in the published Auto top five has a line-level retrieval problem, not a brand-level one. Auto queries are situational — rates, switching, claims, coverage after an incident — and whether a carrier is retrieved for them depends on whether it publishes against those situations rather than against its own product taxonomy. That is a scoped diagnostic with a scoped answer.

Move recommendations from moderate to strong

The metric that decided this ranking is the strong-to-moderate mix inside an unchanged recommendation count, and it is the single cheapest thing on this page to move: it requires no new reach at all. The work sits in the material an answer engine needs in order to commit — explicit statements of fit, eligibility, exclusions, and comparison, written to survive being lifted out of the page at the passage level.

Compete for third-party citation, not only for owned citation

Publishers convert mentions into citations 2.7 times as often as carriers. A citation program that only publishes is competing inside the weaker of the two structures. The higher-leverage target is being the brand a publisher names in the comparison content that answer engines actually reach for, which is a different discipline from producing more owned pages.

8. Data, method, and corrections

All figures are drawn from the Q2 2026 AIVI™ Insurance Benchmark, a study of 134 evaluated carriers across 300 stratified prompts and 1,500 model responses on five engines, producing 4,328 organic brand mentions and 1,742 organic citations. The entity-type table in section 5 covers the 1,348 of those citations that landed on a brand inside the evaluated cohort; the remainder went to sources outside it.

AIVI™ is an organic-only, first-brand-mention composite of share of model (35%), recommendation strength (25%), position weighting (20%), citation influence (15%), and top-three rate (5%), normalized so the category leader scores 100. Recommendation points weight a strong recommendation at 5 and a moderate recommendation at 3. The top-ten machine-readable dataset is published at /api/aivi/insurance/q2-2026/top10.json and the per-brand record at /api/aivi/insurance/q2-2026/brands/allstate.json, both under CC BY 4.0.

The controlling rules behind every number above — prompt-set construction, attribution, the component weights, and the known limitations — are published in the versioned AIVI™ methodology and correction policy. Prompt design follows the generative-engine framing set out in Aggarwal et al., GEO: Generative Engine Optimization (KDD 2024). The published dataset is declared as a schema.org Dataset under CC BY 4.0, and the carrier universe is framed against the National Association of Insurance Commissioners record.

If any figure on this page is wrong, we would like to know. Corrections go to kwalsh@brainpan.ai and are handled under the published correction policy — acknowledged in two business days, determined in ten, and applied to the page with a dated note rather than silently.