Why Your Competitor Shows Up in ChatGPT and You Don't

Someone asked ChatGPT "what's the best tool for X" today, and it named three companies. You weren't one of them. Your competitor was. That single answer is now doing the work a first-page Google ranking used to do, except there's no second page to scroll to and no ads to buy your way onto.

This is not a fluke or a glitch you can wait out. 51% of buyers now start their research in an AI chatbot, up from 29% a year earlier, according to G2. The research starts in the chat, the shortlist forms in the chat, and by the time someone lands on your site they've already decided who the real options are. If the model doesn't know you exist, you're not on the shortlist. You're not even in the room.

The good news: the reason your competitor shows up and you don't is mechanical, not magical. It comes down to a handful of things they did and you didn't. Let's break down what's actually happening and what to do about it.

AI answers are not Google rankings in disguise

The first mistake is assuming your SEO work carries over. It mostly doesn't. Only about 12% of AI-cited URLs overlap with Google's top 10, per Ahrefs. The two systems are pulling from different pools. In fact, the share of Google AI Overview citations coming from Google's own top 10 results fell from 76% to 38% in under a year, also per Ahrefs. The ground is moving fast, and the link-building playbook that earned your rankings is not the thing earning citations.

So what is? Ahrefs studied 75,000 brands and found that YouTube mentions and branded web mentions predict AI visibility far better than Domain Rating does. Translation: the models care less about how authoritative your domain looks and more about how often your brand name shows up in conversations, lists, and videos across the open web. Your competitor isn't winning because their site is stronger. They're winning because they're mentioned more, in the places the model reads.

And it reads a surprisingly narrow set of places. Over 25% of ChatGPT's US citations come from just Wikipedia and Reddit, according to 5W Research. That's two sources accounting for a quarter of everything. If you've never thought about your Reddit footprint or whether you'd survive a Wikipedia notability check, your competitor probably has.

The comparison prompt is where deals are won and lost

Here's the prompt that should keep you up at night: "Compare [your competitor] vs [other competitor]." Notice your name isn't in it. When a buyer types that, the model pulls from comparison pages, review sites, and roundups that already framed the market. If those pages don't mention you, you're invisible at the exact moment a buyer is choosing.

This is why "vs-pages" matter so much. Your competitor likely has a page titled "[Their Product] vs [Yours]" that ranks, gets cited, and frames the comparison on their terms. You can build the same pages, and you should build them for every realistic alternative a buyer would name. These pages do double duty: they get cited in comparison answers, and they let you control the narrative instead of letting a competitor write it for you.

The deeper problem is that AI answers are brutally short. In local search, AI names only three to five businesses, and roughly 1.2% of locations get recommended by ChatGPT, versus 35.9% that show up in Google's local 3-pack, per the available data. Google gives you a long tail. AI gives you a podium. There is no page two. Either you're one of the few names, or you don't exist.

Get into the listicles AI already cites

Stop thinking about pages you own and start thinking about pages that mention you. The single highest-leverage move in GEO is getting onto the "best X" and "top 10 Y" listicles the models already pull from. When ChatGPT recommends tools, it's frequently reading a roundup someone else wrote. If you're on that roundup, you get named. If you're not, you don't.

This is more controllable than it sounds. Yext analyzed 6.8 million citations and found that 86% of the sources AI cites to recommend businesses are ones you can influence. Directories, review platforms, industry roundups, your own structured content. Eighty-six percent. The story that AI visibility is a black box is mostly an excuse. Most of what feeds the answer is sitting in places you can touch.

The plays that work:

• Find the listicles that already rank for your category and get added, updated, or pitched in.
• Build vs-pages for every competitor a buyer might name instead of you.
• Earn mentions on Reddit and in YouTube content, since both punch far above their weight.
• Pursue a Wikipedia presence if you legitimately clear the notability bar.

Review velocity is the signal everyone underrates

Reviews aren't just social proof anymore. They're training data. The models read review platforms to decide who to recommend, and recency matters as much as volume. A steady stream of fresh reviews tells the model you're active, real, and currently relevant. A wall of reviews that all stopped two years ago tells it you're yesterday's answer.

Review velocity, the rate at which new reviews come in, is the lever. Your competitor who shows up in ChatGPT very likely has a more recent, more consistent review flow than you do. This is one of the few GEO inputs you can move in weeks, not months, by simply asking happy customers at the right moment and routing them to the platforms that get cited.

What actually moves AI visibility, ranked

Princeton researchers (KDD 2024) ran a causal study on what changes whether AI recommends you. Treat the magnitudes below as directional, but the direction is clear.

Tactic	Effect on AI visibility	Notes
Add citations to authoritative sources	~+30% (directional)	Link out to credible, recognized references
Add statistics to your content	~+32% (directional)	Specific, sourced numbers beat vague claims
Add expert quotations	~+41% (directional)	Named experts, real attribution
All of the above, on a lower-ranked page	up to +115% (directional)	The underdog gains most
Add an llms.txt file	~zero	No major AI provider honors it

Read that last row twice. llms.txt has roughly zero correlation with citations and no major AI provider honors it. If a vendor is selling you on llms.txt as your AI strategy, they're selling you a placebo. Meanwhile the boring stuff, statistics and quotes and credible citations, is what actually moves the needle, and it moves it most for the smaller players. The largest lift hit lower-ranked pages hardest, which means the gap between you and your competitor is more closeable than it looks.

Where to start

The fastest way to know why you're losing is to see exactly which prompts your competitor wins and which sources the models cite to recommend them. That's a diagnosis, not a guess. You can run a free audit at https://getrecommended.net/audit to see where you currently show up in AI answers, which competitors are taking the slots, and which fixable sources are feeding those answers.

If you want the playbook to act on it, the $39 GEO Action Kit walks through the vs-pages, listicle placements, review-velocity, and content moves above in order of impact, so you're working the highest-impact levers before the ones that do nothing.

Your competitor isn't smarter. They just showed up where the model reads first. You can too, and most of it is within reach this quarter.

FAQ

Why does my competitor appear in ChatGPT when my SEO is better? Because AI citations and Google rankings come from different pools. Only about 12% of AI-cited URLs overlap Google's top 10 (Ahrefs). AI weighs branded mentions, YouTube, reviews, and roundups more than raw domain authority, so a competitor with more mentions can outrank a stronger domain.

How do I get ChatGPT to recommend my business? Get cited where it reads. Yext found 86% of the sources AI uses to recommend businesses are ones you can influence (6.8M citations). Focus on listicle placements, vs-pages, fresh reviews, and adding statistics, expert quotes, and authoritative citations to your content.

Does an llms.txt file help AI visibility? No. llms.txt has roughly zero correlation with citations and no major AI provider honors it. Spend that effort on content credibility signals instead.

What single content change moves AI visibility the most? Adding expert quotations showed the largest effect in Princeton's KDD 2024 study (+41%, directional), followed by statistics (+32%) and authoritative citations (~+30%). Lower-ranked pages gained the most, up to +115%. Treat these magnitudes as directional.

Why is showing up in AI harder than ranking on Google? AI gives you a podium, not a page. In local answers, AI names only three to five businesses, and about 1.2% of locations get recommended by ChatGPT versus 35.9% appearing in Google's local 3-pack. There's no page two to fall back on.