# How do I analyse why ChatGPT recommends our competitors in its answer to a question? > If ChatGPT recommends your competitors, read the answer in five steps: the question, the cited sources, the stated reasons, who else is named, and repetition. Published: 2026-11-07 To see why ChatGPT recommends your competitors for a question, break the answer into five parts: the question itself, the sources the answer cites, the reason given for each recommendation, the other brands named, and what happens when the same question is asked again. GeoMono is a platform that measures how visible brands are in AI answers; this guide shows what you can conclude from a single answer, and where you cannot conclude anything. The short answer: the reason a competitor is recommended is often visible in the sources the answer relies on. Where there are no sources, the reason is not visible either. Keeping those two cases apart is the most important rule of the analysis. ## Step 1: Record the question The first part is the question. The same need, phrased differently, can bring different brands. "Best payroll software" and "payroll software for a small accounting office" are two questions and can bring two different lists of competitors. Record the exact question, the date and the version of ChatGPT you used. Keep both your own name and any competitor's name out of the question. If you ask "What can I use instead of X?", an answer listing X's competitors is exactly what you asked for. Start with the question a buyer would ask before knowing any brands. Note the context too. Earlier questions in the same conversation, personalisation settings and location can all affect the answer. Where you can, ask in a fresh conversation and write down what you held constant. ## Step 2: Did the answer search the web, and what did it cite? According to [OpenAI's help page](https://help.openai.com/en/articles/9237897-searching-the-web-with-chatgpt), ChatGPT can search the web automatically when it judges that a question would benefit from current information. Some answers come with source links; some cite nothing. This is where the analysis splits in two. **If the answer cites sources**, you have the strongest clue. Open every source and check: - Is the competitor named on this page, and how is it described? - Are you named on this page? If not, is this a page that lists brands in your category? - Whose page is it: a comparison site, a forum, a news site, a marketplace, the competitor's own site? - When was it written? An old list may describe a situation that no longer exists. Often a competitor is recommended because it is named on every one of the few pages the answer drew on. If among those pages there is a comparison or list that leaves you out, that is the most concrete finding of the analysis. Being cited is not the same as ranking well. A [2025 Ahrefs analysis](https://ahrefs.com/blog/ai-search-overlap/) found that only a small share of the links AI assistants cite rank in Google's top 10 for the same question. So rather than guessing from your rankings report which page was read, open the links the answer actually shows. **If the answer cites nothing**, you cannot see from outside what it rests on. The model may have recommended the competitor from what it learned earlier. Saying "the competitor was recommended because of this source" would be a guess. Write it down plainly as "no sources cited, reason not visible". ## Step 3: Read the reason given When ChatGPT recommends a brand it usually gives a reason: "easy to use", "suited to small businesses", "offers Turkish-language support", "wide range of integrations". Those reasons show what makes the competitor stand out in a buyer's eyes. Ask two questions of each reason. Is it true, and does it rest on a source? And do you offer the same thing, written somewhere the engine can read it? A feature you have but never state plainly anywhere can look like the competitor's advantage in the answer. Look at the tone as well. Is the competitor named as the leading option, as one of the alternatives, or with a caveat? Being on the same list is not the same as being recommended the same way. ## Step 4: Who else is named alongside you? List every brand named in the answer. Some may not be your direct competitors; the engine may be placing you in a different category. Seeing an enterprise software company listed next to small-business tools is a clue to how the engine classifies you. Make sure each name belongs to the company you think it does. It may be another company with the same name, a product name or a person. A name appearing does not prove that company was named. {{cta}} ## Step 5: Ask again, and count One answer is one sample. [Research published by SparkToro](https://sparktoro.com/blog/new-research-ais-are-highly-inconsistent-when-recommending-brands-or-products-marketers-should-take-care-when-tracking-ai-visibility/) in January 2026 found that AI tools return very different lists of brands when the same question is asked repeatedly. A competitor recommended once is not necessarily recommended every time. Ask the same question several times, in fresh conversations. Then count in how many answers the competitor was named, in how many you were named, and which sources came up again and again. A source cited repeatedly is a much stronger clue than one that appeared once. Where you can, ask the other engines too. Whether a competitor is recommended only by ChatGPT, or also by Gemini, Perplexity, Claude and Google AI Overviews, tells you whether the reason is specific to one engine. ## What the analysis can and cannot tell you This analysis tells you what was observed; it does not prove what caused it. A competitor named on a page that the answer cites is a strong clue, but getting yourself onto that page does not guarantee the answer will change. What you can conclude: - Which pages are read again and again for this question, and whether your name is on them. - The reasons given for recommending the competitor, and whether they are true of you too. - Which category the engine puts you in, and who it names alongside you. What you cannot conclude: - Why an answer that cites nothing recommended that competitor. - That a competitor is "always" recommended, from a single answer. - That a particular change will certainly change the answer. ## How GeoMono runs this analysis GeoMono runs these five steps at scale. Questions are registered before the measurement in a written, versioned declaration and are written without the brand's name. Each question is asked of ChatGPT, Gemini, Perplexity, Claude and Google AI Overviews separately, and every answer is recorded. The Sources and Competitors module collects, by domain and engine by engine, the pages behind the answers that did not name you. Brands named alongside you in the answers are recorded as co-mentions, and each competitor gets its own file: in which questions, on which engine and in what language it was named. Answers whose source list could not be collected are counted separately, never written as "no sources". A co-mention does not say who is ahead, and the module says so plainly. The possible causes on your own side (crawler access, JavaScript, namesakes) are covered under seven headings in [Why doesn't ChatGPT recommend my brand?](/en/blog/chatgpt-markami-neden-onermiyor) ## Frequently asked questions ### ChatGPT recommended a competitor but cited no sources. How do I find out why? You cannot see from outside what an uncited answer rests on. Ask the same question several times; in some repetitions the engine may search the web and cite sources. If none of them do, record it as uncertainty, not as a finding. ### Will appearing in the source that recommends my competitor change the answer? There is no guarantee. A cited source that names the competitor is a strong clue, but it has not been shown that getting onto that page will change the answer. After making the change, ask the same questions again and measure the result. ### Is one question and one answer enough to analyse? No. Engines do not recommend the same brands every time. To see how often a competitor is recommended, you need to ask the same question again and add similar questions. ### The list includes brands that are not our direct competitors. What does that mean? The engine may be placing you in a different category. That can be a sign that what you do, and who you serve, is not stated clearly enough on your site and in the sources written about you. ## Sources - [Searching the web with ChatGPT](https://help.openai.com/en/articles/9237897-searching-the-web-with-chatgpt), OpenAI Help Center, accessed 7 October 2026 - [AIs are highly inconsistent when recommending brands or products](https://sparktoro.com/blog/new-research-ais-are-highly-inconsistent-when-recommending-brands-or-products-marketers-should-take-care-when-tracking-ai-visibility/), SparkToro, 28 January 2026 - [Only 12% of AI Cited URLs Rank in Google's Top 10 for the Original Prompt](https://ahrefs.com/blog/ai-search-overlap/), Ahrefs, 11 August 2025 - How we measure: the GeoMono method, 2026-11-07: published method; weights, thresholds and figures are read from the code that computes the score (https://geomono.com/en/yontem)