ChatGPT recommends your competitors but not you: how to close the gap

When ChatGPT recommends your competitors and not you, the fix is a list: for each buyer question, record which competitors each engine names, open the sources the answer cited, and check whether you can get onto those sources. They are usually "best of" articles, reviews and directories. Get listed, reviewed or included there, publish an honest alternatives page on your own site, and re-run the same questions monthly.

The recommendation is not random taste. It is assembled from pages the engine found, and those pages name your competitors and not you. Below is how to find them and what to do about each kind, checked in September 2026.

Why does ChatGPT recommend your competitors instead of you?

Because they are described on more of the pages it reads. ChatGPT search rewrites a question into one or more search queries, sends them to its search partners, reviews the results and writes an answer from them, according to OpenAI. If the pages that rank for those queries list your competitors, the answer lists them too.

Google's AI features work the same way. Google says AI Overviews and AI Mode "may use a 'query fan-out' technique, issuing multiple related searches across subtopics and data sources," and its guide describes retrieval from "our Search index." An answer to "best invoicing app for designers" is built from the pages that rank for that question and its variations.

The research on who gets named points the same way. In Ahrefs' 75,000-brand study, branded web mentions correlated with AI visibility at 0.66 to 0.71, and the three assistants largely named the same brands, which Ahrefs summed up as "mostly mentioning Nike, Apple, and Amazon." Established competitors win because the web talks about them more. Ahrefs also found ChatGPT the least gated by established authority, which makes it the most realistic place for a smaller product to start closing the gap.

How do you build a competitor gap list?

Run your buyer questions in each engine, record who is named and which pages are cited, then sort those pages by type and by whether you can get onto them. The result is a to-do list of specific URLs, not a general feeling that "AI prefers the big players."

1. Write 10 questions a buyer asks before they know your name: best-of, alternatives, use case, comparison.
2. Run each one in ChatGPT with Search on, Perplexity, Google AI Overviews, AI Mode, Gemini, Claude and Copilot.
3. For each answer, record the brands named, in order, and whether you are among them.
4. Open the Sources panel or the linked citations and copy every URL.
5. Label each URL by page type: listicle, review, directory, comparison on a competitor's site, community thread, video, encyclopedia.
6. For each URL, answer one question: can you realistically get your product onto this page?
7. Turn every "yes" into a named action with an owner and a date.

Use one row per prompt and engine. Ten questions across seven engines is 70 rows, and the same URLs will appear again and again, which tells you which pages matter most.

What does a competitor gap list look like?

It is a table of prompts, winners and sources, with an action on every source you can influence. The example below is for a hypothetical invoicing tool. The actions differ by page type, which is why labelling the source matters more than counting how often a competitor is named.

The third row is common and easy to misread. When the cited page is a competitor's own site, you cannot get onto it; you can only publish a page of your own that answers the same question more directly.

How do you get onto the sources that name your competitors?

Work through them by type, starting with the pages that appear for the most prompts. Directories are the fastest, listicles take a pitch, reviews take a product someone can try, and community threads take an honest answer. Profound's citation data shows why the list is broad: ChatGPT's most cited domain was Wikipedia, while Reddit led Perplexity and AI Overviews.

What to send a listicle author, and how to make the pitch worth their time, is in how to get listed in "best of" articles. Why engines trust these third-party pages over your homepage is covered in why AI answers cite review sites and directories.

Should you publish a "competitor alternatives" page?

Yes, if it is honest. A page titled "[Competitor] alternatives" answers a question buyers ask and engines search for, and it is one of the few gap pages you control. It works when it treats the competitor fairly, names who each option suits, and gives specifics a reader can check.

Google's guidance on reviews describes what makes such a page useful: "Explain what sets something apart from its competitors," "Cover comparable things to consider, or explain which might be best for certain uses or circumstances," and "Discuss the benefits and drawbacks of something, based on your own original research." When you call something the best for a purpose, it asks you to "include why you consider it the best, with first-hand supporting evidence."

Build one page per main competitor, not one per phrasing. Google's generative AI guide warns that creating separate content "for every possible variation" of a query "primarily to manipulate rankings or generative AI responses" violates its scaled content abuse spam policy. How to structure these pages is covered in alternatives pages for SEO.

A good alternatives page includes:

1. The competitor's real strengths, stated plainly.
2. A comparison table with prices, limits and features, dated.
3. Who should choose the competitor and who should choose you.
4. Your one-sentence description and pricing, matching the rest of your site.

What if engines are not reading your site at all?

Check this before chasing sources. If you are missing even when a buyer asks about your exact category, and engines cannot describe you when asked by name, the problem may be access rather than reputation. An engine that cannot fetch or read your pages cannot cite them, however many listicles you appear in.

Run four checks:

1. Robots.txt: OpenAI says sites that block OAI-SearchBot "will not be shown in ChatGPT search answers." Make sure OAI-SearchBot, PerplexityBot and Claude-SearchBot are allowed, and that a wildcard rule does not catch them.
2. Firewall: a bot-protection setting can challenge these crawlers before robots.txt is ever read. Allow the IP ranges each operator publishes.
3. Rendering: Vercel's measurements found that OpenAI's, Anthropic's and Perplexity's crawlers do not render JavaScript. Fetch your key pages with curl and confirm the text is in the response.
4. Bing: ChatGPT search partners with Microsoft among others, so confirm your key pages are in Bing's index with a site: search on bing.com.

How do you know the gap is closing?

Re-run the same prompts every month and track how often you are named per prompt, not where you appear in the list. SparkToro's study of 2,961 runs found less than a 1 in 100 chance that ChatGPT or Google's AI gives the same list of brands twice in 100 runs, while how often a brand appeared across many runs was far more stable.

Watch three numbers each month: the share of answers that name you, the number of cited sources you are now on, and your share of voice against the competitors in your sheet. For Copilot, Bing Webmaster Tools' AI Performance report adds Citation Share, "the percentage of citations attributed to your site out of all citations shown for a specific grounding query." The full measurement method is in how to measure AI visibility.

Get the competitor gap read for you

LaunchScaler's full audit runs this analysis on your domain. It puts 10 buyer questions to 7 answer engines, reads all 70 answers, and reports which competitors each answer names and on which prompts, your citation share against them, and which buyer questions leave you out. It costs $19 once for your domain and unlocks on your report after the free scan. Run the free scan first, then open the full audit.