Why ChatGPT and Gemini Don't Recommend Your Business: Causes and Action Plan

You ask ChatGPT or Gemini about your category and they name three competitors. Your brand isn't there. It isn't a system error or bad luck: AI assistants recommend the brands they find clear, consistent, easy-to-cite evidence about. This guide covers the most common causes, how to confirm the problem is real, and a five-step plan to close the gap.

First, confirm the problem is real

A single question from your own account isn't enough. The answer changes with the account, the memory of earlier conversations, the location, and even between two runs in a row. For a first check, use a temporary chat, ask 10 to 15 questions like the ones a customer in your category would type ("best [product] for [use case] in [country]", "X vs Y" comparisons, "is X reliable?") and repeat them in ChatGPT and in Gemini separately. Note whether you appear, in what position, and who they name instead of you: that last detail is the most useful, because it tells you which content you're competing against.

If you don't appear anywhere, rule out the technical side first: whether bots can read your site at all (robots.txt, content that depends on JavaScript, pages with noindex). We walk through it in 5 technical reasons ChatGPT can't find your brand.

The 6 most common causes

1. 01
   The model doesn't recognize you as an entity: If your name, category and country aren't consistent across your site, LinkedIn, directories, press and, where it applies, Wikidata, the AI doesn't really know who you are or what you do, and may confuse you with another company. It's the cause people skip most. Guide: how to create your brand's entity on Wikidata.
2. 02
   Your content can't be cited: Marketing paragraphs without a single sentence that answers the question don't survive synthesis. Models cite clear definitions, lists, comparisons, questions with direct answers and data with a source.
3. 03
   Nobody talks about you where the model looks: Recommendations lean on third-party sources: trade media, comparisons, directories, review platforms, forums. If your competitor has far more independent mentions, it wins the recommendation even if your site is better. A mention doesn't need a link.
4. 04
   You have no comparison or positioning content: "X vs Y" and "best for [use case]" questions are answered by pages that compare and say who each option is for. If your competitor has them and you don't, the model uses them to place that competitor in the market, and not you.
5. 05
   Your story is inconsistent: If your website says one thing, your LinkedIn another and a press note a third, the model can't work out what you're best at. Independent sources that agree on your category and use case strengthen the recommendation; contradicting ones weaken it.
6. 06
   You don't measure, so you don't find out: Most brands discover the problem by accident: when a customer mentions it or someone on the team tries it. Without systematic measurement you don't know which queries you appear in or when a competitor displaces you.

When your competitor shows up and you don't

It's the most frequent case and the most actionable one, because the gap can be measured. Run the same prompts for your brand and your competitor's and compare mention rate, average position and the kind of question each one wins. Then look at which sources the engine cites when it answers: that's the map of where you need to appear. In Lumen AI you see it in the Authority Sources panel and in competitor comparison.

Being bigger isn't what decides it. A niche specialist can win specific queries if it has clear content, a well-defined entity and consistent mentions, even when the competitor is a generalist.

ChatGPT and Gemini don't answer the same way

It's common to show up in one assistant and not the other: each works from its own sources, its own update cycle and its own criteria, and answers change over time. That's why you measure them separately. On Google's side, its AI features (AI Overviews and AI Mode) rely on the same ranking and quality systems as Search: to appear as a supporting link, a page has to be indexed and eligible to show with a snippet, with no extra technical requirements or special markup. According to that same Google guide, Search doesn't use llms.txt files either: they may help with other services, but don't treat them as your main lever.

5-step action plan

1. 01
   Audit your baseline: Run 10 to 15 prompts from your category in ChatGPT and Gemini, for your brand and two or three competitors, and save the answers. It's your starting point.
2. 02
   Write citable content: For every question you don't appear in, publish a page that answers it in the first lines, with data and sources, and names your brand clearly. Add honest comparisons ("X vs Y", "best for...").
3. 03
   Earn mentions in the sources the engine already cites: Prioritize the domains that show up as sources in answers for your category: comparisons, trade press, directories, reviews. A few mentions in the right sources weigh more than many anywhere else.
4. 04
   Unify your entity: One or two sentences of description, identical on your site, LinkedIn, directories and review profiles; Google Business Profile if you have a local presence; Wikidata if you meet its criteria.
5. 05
   Measure week by week and adjust: Repeat the same prompts every week and log mention rate, position and who they name instead of you. Put the effort where the gap is widest.

How long until you see a change?

There's no guaranteed timeline, and it's worth distrusting anyone who promises one: it depends on your starting point, how competitive the category is and when each engine refreshes its sources. What you can do is measure every week with the same prompts. Mention rate is usually the first metric to move; position improves later. What each number means is in how to interpret GEO metrics.