For every time AI names your business, it names a competitor at least 3.7 times

Every business owner worries their AI-visibility score is low. But a score is abstract. The concrete, gut-punch version of the question is: when a potential customer asks an AI engine to recommend a business like yours, whose name comes up? We ran the count across 173 completed AskedAbout checks — 64 med spas, 42 personal-injury law firms, 40 dentists and 27 home-services companies, 1,024 answers from ChatGPT and Perplexity — tallying every mention of the audited business against every mention of a rival.

Finding 1: AI names a competitor at least 3.7 times for every time it names you

Across all 173 businesses, the engines named the audited business 470 times and named a competitor 1,748 times in the same set of answers. That's a 3.7-to-1 ratio — in the answers that are supposed to be about your business, a rival's name appears nearly four times as often as your own. And this is a lower bound: our audit records only the top five rivals per business, so every mention beyond the fifth competitor goes uncounted. The real ratio is worse.

Finding 2: It's the same ratio in every industry — this isn't a med-spa quirk

You might expect crowded categories to skew the number. They don't. The rival-to-own ratio is almost identical across four very different verticals:

When a pattern holds this tightly across med spas, law firms, dentists and HVAC companies alike, it's not about your industry being unusually competitive. It's structural: the engines answer "who's good near me?" by assembling a shortlist of businesses, and most businesses are simply not on their own shortlist very often.

Finding 3: 43% of businesses are out-named by a single rival

The aggregate ratio is one thing; the head-to-head is sharper. In 75 of the 173 audits (43%), one specific competitor was named more times than the audited business itself. Not the whole field out-mentioning you — a single, nameable rival, being recommended ahead of you, in the answers your own prospects are reading. For nearly half of these businesses there is a concrete competitor winning the AI conversation, and most of them have no idea who it is.

The flip side of the same coin: a business was named in only 45% of the answers about its own city and category. More often than not, the engine builds its recommendation and the business isn't in it at all.

Finding 4: But no single rival owns the answer either

Here's the part that turns this from bad news into an opportunity. Those 1,748 competitor mentions are spread across 633 distinct businesses. The single most-named competitor in the entire dataset holds just 0.9% of all mentions (15 of 1,748). There is no incumbent, no business that has "won" AI recommendations the way a site can own the #1 organic result for a keyword. The rivals beating you are, individually, barely ahead — the lead is measured in a mention or two, not in years of entrenched authority.

So the picture is: AI is recommending a competitor instead of you far more often than it recommends you, but the competitor it recommends is different almost every time and none of them are far in front. That's a closeable gap, not a lost war.

What this means if you run — or market — a local business

For an owner: the useful move is not to obsess over your score, it's to find the specific competitor being named ahead of you and the specific questions where it happens — then work the off-site sources the engines actually read (review platforms, local directories, "best-of" roundups; see where AI looks before it recommends a local business). You don't have to out-rank the whole field. You have to close a one-or-two-mention gap against a rival who isn't defending it.

For an agency or GEO consultant, this is the pitch, quantified: your client is being out-named nearly 4-to-1, there's a specific competitor doing it, and nobody owns the category — so the ground is winnable and the work (seeding and cleaning the third-party corpus) is ongoing, which is exactly what a retainer is for. Walking in with "here's the competitor AI names instead of you, and here's how far ahead they actually are" beats another audit-score slide. Reselling this is what the white-label Agency 5-pack is for — five of these audits under your logo for $249.

Either way the first step is the count itself: which competitors AI is naming instead of a given business, on which questions, per engine — which is exactly what an AskedAbout audit reports.

Methodology

173 completed AskedAbout free AI-visibility checks across four local-service verticals — 64 med spas, 42 personal-injury law firms, 40 dentists, 27 home-services companies (June–July 2026), 1,024 parsed answers. Each check submits buyer-intent, category-and-location questions to the two engines in the free check — OpenAI's ChatGPT search model and Perplexity — and counts mentions of the audited business and of named competitors. "Own mentions" is the number of answers naming the audited business; "competitor mentions" sums the counts of every rival named in the same answers. The per-audit competitor tally is capped at the top five rivals, so total competitor mentions are undercounted and the 3.7:1 ratio is a lower bound. Corrected 2026-07-21: an earlier version of this page counted three engineering QA rows as businesses; `audit_cut` now excludes them by exact name, which moved the ratio from 3.8 to 3.7 and n from 176 to 173. Synthetic smoke-test rows are excluded by business name via our canonical `audit_cut` script — never by test flags, which our own prospect audits also carry. n and per-vertical splits are stated so you can weight them; we re-publish as the corpus grows.

More from the same corpus: the full AI-visibility benchmark across 152 local businesses, where AI looks before it recommends a local business, and — written for owners — why AI recommends your competitors. Reselling this to clients? The white-label Agency 5-pack puts five of these audits under your logo for $249.