Getting mentioned by ChatGPT comes down to one thing: your site has to carry reliable, citable signals that match the buyer questions you care about. The fastest way to check is to look at what ChatGPT actually says sentence by sentence, then fix what’s missing in your public materials.
Check what ChatGPT is already saying, sentence by sentence
If you want to know why you’re not mentioned, start with a direct check. According to Yuzu’s measurement approach, the job is to see whether your brand gets named when buyers ask for a recommendation, and whether the answer includes a citation that links to your site.
The point is simple: if your brand isn’t named, or it’s named in the wrong sentence, you need to see the exact text that was used. Yuzu measures (1) whether buyer questions in your category get your brand named in AI answers, (2) the ranked field, meaning which brands the engines recommend instead, and (3) citations, meaning when an answer links your site, not just names it.
If you only skim the final answer, you’ll miss the actual failure mode. Sometimes it names someone else. Sometimes it doesn’t name anyone. Sometimes it names you but doesn’t cite you. If you have ever copied an answer into a doc and highlighted the sentence that mattered, that is the level to work at here.
For a manual workflow you can run today, use the existing checklist in: How to find out what ChatGPT and Perplexity say about your brand. That post focuses on checking by hand and separating a real measurement from a made-up number.
Map the buyer questions you want to win
You don’t get mentioned because you “have good content.” You get mentioned when your public pages match the questions buyers ask answer engines.
According to Yuzu, you add your domain, then it derives the questions buyers ask from your own homepage, puts those questions to the answer engines, and records who was named. In other words, it turns “mentions” into a measurable outcome tied to specific buyer-style prompts.
The practical translation is: decide what recommendation questions matter in your niche, then ensure your website has the answers in a form that can be reused as citations.
If you’re coaching, advising, or counseling, the “buyer question” might be something like “Who would you recommend for X?” If you’re a lawyer or finance professional, it might be “Which firm handles X?” The format matters because models tend to reuse whatever reads like a recommendation signal they can point to.
Confirm whether you’re missing names, citations, or both
A lot of people fix the wrong thing because they don’t separate naming from citation. Your goal is the evidence trail as well as the brand string in the response.
Yuzu distinguishes:
- Whether you’re named in ChatGPT and other answer engines for buyer recommendation questions.
- The ranked field: which brands the engines recommend instead.
- Citations: when an answer links your site, not just names it.
That split changes your next move. If you’re named but not cited, you might need to reorganize the content so the relevant sentence can be referenced directly from your page. If you’re not named at all, you likely need to align your public pages with the buyer questions that trigger recommendations. In practice, that can be as basic as checking the service page you expected to be used and seeing whether it actually answers the question the model was asked.
Use a failure driven debug loop
My rule here is simple: treat “not mentioned by ChatGPT” as a debugging problem.
Start by checking what ChatGPT says. Then compare what it uses to what you’ve published. If ChatGPT does not mention your brand, the most common reason is that the model doesn’t find reliable, citable signals that match what buyers ask.
If you want the most direct version of that idea, see: Why is my brand not mentioned in ChatGPT?. That post walks through checking what ChatGPT says sentence by sentence and fixing what the answers are missing.
To make this concrete, use the table below to decide what to change based on what you observe.
| What you see in ChatGPT’s response | What it usually means | What to do next | |---|---|---| | Your brand is not named | The engine didn’t find matching signals for the buyer recommendation question | Check the exact questions that trigger recommendations, then align your public pages to answer them clearly (and in a way that can be cited). | | Your brand is named, but the answer doesn’t link your site | The engine may be using your name without a citeable link to your content | Find the referenced sentence in your own pages and adjust so that the relevant answer can be linked from your site. | | A competitor is named instead | The ranked field points to other brands with stronger match signals for that question | Identify the competitor who gets named, then compare the type of content and coverage that appears to be used for the recommendation. |
Your fixes should be driven by what the response actually includes. If a competitor keeps showing up for one service line, start with that exact prompt and page pair before you touch the rest of the site.
Publish content that can be cited
“It sounds right to a human” isn’t a measurement. You need content that creates usable, citable signals.
According to Yuzu’s accuracy notes, the product is designed so every number links to the answer text it was read from, so a reading can be checked and verified. That is the standard you should apply to your own visibility efforts.
Practical guidance based on that standard:
- Make sure your site contains the plain-language answers to the buyer-style recommendation questions you’re trying to win.
- Ensure those answers live on pages that plausibly match the question the model is answering.
- Reduce ambiguity in the part you want to be cited, because the model will reuse sentences it can map to the prompt.
If you’re a consultant, therapist, lawyer, or adviser, add the details you want quoted. Don’t hide them behind generic taglines. If a prospective client asks who handles a specific issue, the sentence you want reused should already be sitting on the page.
Check Google Search Console for discovery
Even if your immediate target is “mentioned by ChatGPT,” discovery signals from search still matter in the measurement loop.
According to Yuzu, it includes search visibility alongside measurement from Google Search Console, and it also tracks search console metrics like impressions and clicks, plus index status. The workflow is: measure what’s indexed and visible, then ask the answer engines, then compare what gets named and cited.
So if you’re not appearing at all, don’t only check the chat experience. Check that your site has the basic discovery footprint that lets a system find it and then cite it. A page that is not indexed is not going to help much, even if the copy on it is good.
Run it locally as a repeatable checklist
Here’s a short checklist you can run per brand or per offer:
- Step 1: Pick 10 buyer-style recommendation questions you want to win.
- Step 2: Ask ChatGPT each question and copy the exact response sentences that include brand names and any linked citations.
- Step 3: Identify the failure mode: not named, named without citation, or competitor named instead.
- Step 4: Update the specific pages that should contain the answer sentence you want reused.
- Step 5: Re-check the same prompts, and keep only changes that move the sentence-level output.
This is the same idea as the existing Yuzu blog approach: verify by hand, sentence by sentence, and distinguish a real measurement from a made-up number.