Does brand size matter in AI search?

I compared five major US car insurance brands through 2,600 prompts and 200,000+ brand-source mentions in Semrush One.

Here’s the link to the full post.

GEICO stood out as an anomaly.

Despite having an audience footprint around 50% smaller than Allstate, it appeared in almost 2x as many AI answers in the sample.

So I looked at why.

Not to see who wins overall. But to understand why a brand with a smaller estimated AI-search footprint shows up more often than another, when the obvious signals don’t fully explain it.

The dataset covered AI Visibility, audience estimates, mentions, prompt-level appearances, topic clusters, and source-domain data across ChatGPT, Gemini, Google AI Mode, and Google AI Overviews.

The main thing I looked at was the gap between footprint and mentions.

Allstate had the largest estimated audience footprint in the dataset. GEICO had the smallest. But GEICO still appeared more often across several AI visibility signals.

GEICO had:

A quick caveat.

Audience in Semrush One is not customer count, market share, or website traffic. It’s an estimate of the query audience across topics where the brand appears.

The topic data gave more insights.

GEICO led in:

Progressive led in more technical coverage topics like liability, gap insurance, non-owner insurance, and comprehensive coverage.

So GEICO isn’t being shown more everywhere. It’s being shown more where the topics are tied to affordability, discounts, savings, and quotes.

So why is GEICO being connected to those topics more clearly than Allstate?

GEICO has a dedicated cheap car insurance page that is built around the same language where it overperformed.

The page also answers related questions directly: how to lower rates, what affects price, whether minimum coverage is enough, which cars are cheaper to insure, and what discounts can reduce the cost.

That is useful for AI search because it gives the system clear answer fragments.

Allstate also has a cheap car insurance page. But the message is broader. It includes affordability, but also premium protection, 24/7 service, Drivewise, broader coverage, and protection messaging.

That may work for users comparing coverage quality.

But for cheap-car-insurance queries, GEICO’s page is more directly aligned with answer extraction.

GEICO doesn’t just have one cheap car insurance page.

It has dedicated pages:

That gives AI more entry points.

These aren’t random content assets either, they all point back to the same commercial meaning:

GEICO is connected to savings-led car insurance.

Allstate has useful content too, especially around Drivewise, safe driving, claims, protection, and coverage. But that builds a different association.

Allstate is more connected to protection and safe-driving value.

GEICO is more connected to lower-cost insurance and discounts.

This isn’t about GEICO being a bigger or more reputable brand.

Allstate is a major national insurer. It has strong advertising equity. Its Mayhem campaign is highly memorable and built around risk, chaos, protection, and making sure people are properly covered.

That’s a strong brand position.

But it’s not the same position GEICO owns.

GEICO’s public message has been price-led for years. Its marketing has focused on savings, quick quotes, competitive pricing, and direct insurance.

That gives the two brands different meanings:

GEICO = savings-led car insurance.

Allstate = protection-led insurance.

But AI search isn’t just asking which brand is strong overall. It’s trying to answer specific questions.

For affordable car insurance, discounts, savings, and quote-led topics, GEICO’s meaning is easier to connect to the answer.

AI answers are not built only from brand websites.

In the data, major review and comparison publishers were prominent source domains:

NerdWallet.
Bankrate.
MarketWatch.
Forbes.
The Zebra.

Those sources often reinforce the same GEICO association.

NerdWallet names GEICO as ‘Best ways to save.’ 

The Zebra’s GEICO vs Allstate comparison chooses GEICO and supports that with pricing and score comparisons.

Across the source-domain data, State Farm had the highest absolute source mentions. But when normalized against estimated audience size, GEICO was strongest.

GEICO had 42,007 review-source mentions, which worked out at 170.3 mentions per 1M audience.

Allstate had 29,607 review-source mentions, or 60.5 per 1M audience.

That supports the same pattern.

GEICO isn’t always the absolute leader. But relative to its footprint, it gets more value from review and comparison sources.

Comparison pages are the type of content AI systems can reuse.

Evertune found that listicles accounted for half of unique URLs cited in its dataset, and 63% of citations across nearly 400 million citations pointed to listicles.

That doesn’t mean every listicle is valuable. But it does show why third-party comparison content matters.

If those pages keep framing GEICO around savings, cheaper rates, and discounts, they strengthen the same association GEICO is already building on its own site.

The takeaway is to look at what answer your brand is becoming known for.

AI visibility is not one score. It changes by topic, platform, source, and intent.

Marketers need to look at AI search in terms of:

Because the brands that appear most often are not always the largest, they are often the brands most clearly connected to the topic and answer.