ChatGPT recommends the brands it can find, fetch, and quote for that exact prompt. If competitors have clearer product data, useful category pages, and passages that answer the use case, they get named and you get skipped.

To fix it, run the same buyer prompts in ChatGPT each week, track mentions and citations, then fix the catalogue and pages that failed. Then test again.

How do I measure visibility in ChatGPT results?

You measure visibility in ChatGPT results by running the same set of buyer prompts every week and counting mentions and citations. One-off checks do not count, because answers change by wording.

Use a fixed list so the numbers mean something:

  1. Pick 20 to 30 prompts buyers actually type, from broad use case to exact product need.
  2. Run them clean in ChatGPT and save the full answer and any citations.
  3. Score each run: were you named, cited, or skipped, and which competitors were named?
  4. Repeat weekly and track share over time, not one answer.

That 30-prompt protocol is the simplest honest score. groas plugs into the AI assistant you already use, so you can ask in plain English and work from your real accounts with no dashboard to learn.

How do I check if my brand appears in ChatGPT?

You check if your brand appears in ChatGPT by asking it the way a buyer would, not by asking about your brand by name. Branded questions always find you. Buyer questions show if you are visible.

Start with five to ten prompts from your exact use case. Ask for options, best tools, and what to buy for that job. Run each prompt fresh, note if you are named, linked, or missing, and note who is named instead.

Run the same prompts in new chats and with small wording changes. If you never appear on unbranded prompts, you have a retrieval problem, not a wording problem. That gap is what the fixed buyer prompts and competitor share check is built to catch, and it tells you which use cases to fix first.

Why does ChatGPT recommend my competitors and not me?

ChatGPT recommends your competitors and not you because for that prompt they were easier to retrieve and quote. It is not personal. Their pages gave a clearer answer to that exact need.

Most skips come down to four gaps:

  • Exact use-case pages — Competitors have a page that answers that job in plain words, yours does not.
  • Specific product data — Their catalogue, specs, and feeds make the product easy to match and cite.
  • Useful category proof — Their category pages group options, prices, and proof in one fetchable place.
  • Quotable passages — Their pages have short answers ChatGPT can lift, yours bury the answer.

Start with the catalogue, not the blog. Add specific product data, clean feeds, and useful category pages, then build content around them. If ChatGPT cannot fetch a clear passage, it picks someone it can.

You figure it out by rebuilding that exact prompt and following the fetch. Run the prompt again, save who was cited, then check what those cited pages have that yours does not.

Compare the winners to your page side by side. Look for a direct answer to the use case in the first screen, concrete details like fit, price, and limits, and proof a buyer would trust. Then check if your page can even be fetched for live answers. A server-log test shows which pages ChatGPT actually fetches, separate from crawler traffic and spoofed bots.

Fix one page for that one prompt, make the answer quotable, and re-test that prompt weekly for six weeks. Change what earns citations on your own queries, not what a checklist says. One fixed prompt beats ten rewritten blogs.

What prompts lead to citations in ChatGPT vs. Perplexity?

Prompts that lead to citations are prompts that trigger retrieval. A citation starts with a prompt that needs fresh sources, then search, fetch, and selection. No retrieval, no citation.

In general, specific buying prompts with constraints get cited more than vague ideas. “Best tool for X under Y for Z team” forces a lookup. “Tell me about X” can be answered from memory with no links.

ChatGPT and Perplexity cite different pages for the same prompt because they search, fetch, and pick differently. Follow those three steps to see why one fetched your category page and the other fetched a competitor. Target buyer prompts with clear buying intent, write passage-first answers that quote cleanly, and test both assistants on the same list.

Start free and ask in plain English.