Why does ChatGPT recommend your competitors and not your business?

Why does ChatGPT name your competitors instead of you?

ChatGPT names your competitors because it can find and quote them for the exact question that was asked, not because it has judged them better businesses. There is no editor behind the answer and no directory it consults. The model assembles a recommendation from pages it has read, and if those pages mention your competitors and not you, the answer follows.

The second thing to know is that the answer you saw is not fixed. Rand Fishkin of SparkToro found that when the same prompt is run repeatedly, there is less than a 1 in 100 chance that any two answers name the same list of brands, and eMarketer reports that between 40 and 60 percent of cited sources change from month to month across ChatGPT and Google AI Mode. The competitor who beat you today may be gone next month, which is exactly why a single check tells you almost nothing.

What is ChatGPT actually reading when it recommends a business?

When ChatGPT recommends a business it draws on two layers: what its model absorbed during training, and what its search feature fetches live when you ask. The live layer matters more for recommendations, because the model pulls candidate pages from a search index and then quotes the ones that answer cleanly.

Those candidates are rarely homepages. They tend to be comparison posts, industry roundups, review profiles, and pages that answer a specific buyer question directly. If the only page about your business is a homepage that says you are passionate about excellence, there is nothing quotable, and the engine moves on to a competitor who gave it something to quote.

How do you find out which buyer questions trigger those recommendations?

Start by writing down what your buyers actually ask an assistant, which is rarely the keyword you would type into Google. People ask full questions: best option for a situation, alternatives to a brand they already know, whether a type of service is worth it for a business like theirs. Aim for five to ten of these in your customers' own words.

This matters because recommendations are question-shaped. You might be invisible for "best accountant in Leeds" and quotable for "accountant for a cafe that runs on Square." Testing the wrong questions produces the wrong panic. If you want the background on why phrasing matters so much, what AEO is and why it matters covers it.

How do you check who gets named across ChatGPT, Claude, Gemini, Perplexity, and AI Overviews?

You check by asking every engine the same questions and recording the results the same way. Ask each of your buyer questions in ChatGPT, Claude, Gemini, Perplexity, and a Google search that triggers an AI Overview, and log four things for each answer. Done by hand for five questions, this takes about 20 minutes.

Record	Example
Question asked	"best bookkeeping service for small cafes"
Who was named, in what order	Competitor first, you absent
Which sources were cited	Two roundup posts, one review site
Engine and date	ChatGPT, 31 August

That log is the diagnosis. It shows whether you are losing everywhere or only on one engine, whether the same third-party roundup keeps feeding every answer, and whether the picture holds when you re-check in two weeks. This is also the tedious part, which is why CiteRank runs it on a schedule. It finds the questions your buyers ask, checks all five engines, and keeps a dated record of every check: who was named, in what order, which sources were cited, and an excerpt of the answer.

What are your competitors' pages doing that yours are not?

Open the cited sources from your log and the pattern is usually plain: those pages answer the question near the top, in sentences a model can lift word for word. They use headings phrased the way buyers ask, state who the business is for, and often appear on third-party lists as well as their own site.

Compare that with your own pages honestly. Most businesses that get skipped have the substance but bury it, spreading the answer across an About page, a Services page, and a brochure PDF that no engine will ever quote. The gap is usually structure, not merit.

Can you pay or register your way into the answer?

No. There is no registry to join, no form to submit your business, and no way to buy a place inside the organic answer. Sponsored results are appearing in some assistants, but they are marked as ads, and the recommendation itself is not for sale.

Treat anyone offering guaranteed placement in ChatGPT with suspicion, because they are selling control of a system that reshuffles its own sources monthly. What you can influence is the input: being present, quotable, and consistent in the pages these engines read. Even then, an engine can only quote pages its search layer has already indexed, so changes take time to surface, and nobody honest will promise you a date.

What should you fix first this week?

Fix the single biggest gap your log surfaced, and nothing else yet. If your site blocks AI crawlers, that comes first, because nothing else works until they can read you. If the same roundup post feeds every answer that excludes you, getting listed there is probably worth more than ten new blog posts. If your pages simply never answer the buyer question directly, rewrite one page this week so it does.

Then put a repeat check in your calendar for two weeks out, because a diagnosis without a follow-up reading is just a bad day. When you are ready to work on the fix side, the playbook in how to get ChatGPT to cite your website picks up where this diagnosis leaves off.