AI visibility
Why Does ChatGPT Recommend My Competitor Instead of Me?
A potential customer asked ChatGPT who does HVAC installation in their area. Your competitor’s name came back — business name, phone number, the works. Not yours.
You have a website. You’ve been in business for eleven years. You’re listed on Google. So why did the AI skip you?
The answer is not what most people expect, and it has nothing to do with how your site is built.
Why AI recommends your competitor — and it’s not about rankings
AI engines don’t rank businesses the way Google does. They recite what their training data already established.
When a large language model answers “who does HVAC installation near me,” it’s drawing on patterns in its training corpus: how often your business appeared in sources it has already read, which sources those were, and how much authority those sources carry. The answer it gives is less like a search result and more like a journalist citing the businesses that come up most often in trusted reporting.
The mechanism matters because it changes what fixes the problem.
If your business barely exists in the sources AI engines have indexed — directories, review platforms, local news, industry sites — you won’t show up, regardless of how good your website copy is, how many Google reviews you have, or whether you’ve added an llms.txt file to your root directory.
What actually predicts AI citations — the data
Ahrefs studied 75,000 brands and measured what actually correlates with AI citation share. The findings:
- Brand mentions in trusted web sources (directories, review platforms, local and industry news): correlation of 0.656–0.709 with AI citation share
- YouTube brand mentions: correlation of 0.737 — the single strongest predictor in the study
- Traditional backlinks: weak correlation — the metric most SEO tools optimize for barely registers
This is the gap between your business and your competitor made legible. If they have a Yelp listing, a Nextdoor mention, a TripAdvisor entry, and they’ve been tagged in a couple of local news pieces — and you have none of those — the AI has simply seen them more often, in more places it trusts.
That’s a citation-footprint gap, not a ranking gap. And it’s a different thing to fix.
What this looks like in practice
In one local business audit we ran, we asked AI engines — ChatGPT, Claude, Gemini, and Perplexity — to name service providers in the client’s vertical. The engine that named two competitors didn’t just give their business names. It surfaced their phone numbers.
That information didn’t come from those competitors’ websites. It came from the directories and review platforms that had indexed it — aggregators the AI had read and trusted enough to recite verbatim.
The client’s phone number appeared nowhere in any response. Not because their number wasn’t findable. Because the sources the engine trusted most hadn’t indexed it.
Why the problem is getting harder to ignore
The AI overview share of the top-10 citation results dropped from 76% to 38% in eight months, according to Ahrefs tracking data — meaning the AI engines that used to concentrate citations on a handful of well-known domains are now drawing from a wider source set.
That’s actually good news for local and niche businesses: the pool is widening. But it means the businesses that have built a real citation footprint — mentions across directories, reviews, and local sources — are the ones being pulled into those expanded answers. Businesses that haven’t are falling further behind as the volume increases.
What we found when we ran this on ourselves
We ran our own audit — 100 queries across 5 engines, covering the kinds of searches a potential client would actually run. We appeared in 4 of 100 for presence, 0 of 100 for citations.
In the 40 queries where buyers were actively shopping — “best AI visibility audit,” “top GEO consultants,” “who should I hire” — we showed up zero times. OtterlyAI appeared in those same 40 queries 16 times total. Profound appeared 12 times. Semrush 10.
This was with a live website, a solid technical foundation, and years of experience. None of that showed up in the citation count because none of it had translated into mentions in the sources these engines actually read.
The audit told us exactly what was missing and where. That’s what an AI visibility audit does: it maps the gap between what you’ve built and what the engines have actually registered. More on how these gaps appear across different business types in the AI Visibility Library.
→ We scored ourselves 4 of 100 for presence, 0 of 100 for citations. Here’s the full breakdown.
→ Find out where your business stands: get a free AI Visibility Snapshot.
Ahrefs 75,000-brand study and AI citation data, 2024–2025. AI citation behavior changes as models are trained and updated — if you’re reading this more than six months after the publication date above, check for updated research.