Most owners guess whether ChatGPT, Google AI and Perplexity mention their business. You can check in 30 minutes with a notebook or spreadsheet, a browser and your own robots.txt file. This is the same first look behind our not showing up in AI answers page: ask the questions your buyers ask, write down what comes back, and rule out the technical blocks. It will not make you appear in anything. It will tell you which problem you have before you spend money. If you want the measurement side first, read how to measure SEO and AI answers before the click.
Why is it hard to show up in AI answers?
It is hard because nobody controls the answer and no engine has a list you can apply to. Google says there are no additional requirements to appear in AI Overviews or AI Mode beyond being indexed and eligible to show with a snippet, and no special files or markup are needed. So the work is the same SEO work Google recommends for regular Search, and Google adds that meeting the requirements does not guarantee a page will be shown.
The other engines add their own uncertainty. OpenAI says ChatGPT search sometimes partners with other search providers, and that search results and citations can be incomplete, outdated or incorrect. Google says AI Overviews and AI Mode may use different models, so the answers and links they show will vary, and that AI Overviews often do not appear at all. That is why this self-check uses three engines and repeats every question.
When a business is missing, the cause usually falls into one of four groups: AI crawlers are blocked, the pages are not indexed by the search engine the AI draws on, the pages describe the business in marketing language instead of stating answers, or nobody outside your own site mentions you. The steps below test the first two directly.
Minutes 1 to 10: Run the prompts
Ask each engine the questions a buyer would ask before hiring you, and ask each one twice. Use ChatGPT with search turned on (OpenAI says you can type / in the message composer and select Search), Google AI Mode, and Perplexity. Use a private window or a logged-out session; OpenAI says ChatGPT may use saved memories when it rewrites a search query, and signed-out people can still use ChatGPT search. Write five prompts, one of each type:
- Category plus place: "best [what you do] in [your city]".
- Problem: "how do I fix [the problem you solve]".
- Comparison: "[your business] vs [a competitor]".
- Brand: "what does [your business] do, and is it any good".
- Buying criteria: "what should I look for when hiring a [your type of business]".
Run each prompt twice on each engine. That is thirty answers at most, and you only need to skim each one for your name and your link.
Minutes 11 to 15: Write down what you see
Record facts, not impressions. A spreadsheet with one row per answer is enough, with these columns:
- The exact prompt, the engine, the date and the run number.
- Whether you were named, linked as a source, or absent.
- Which businesses and which pages were named or cited instead of you.
- Anything stated about you that is wrong, such as an old address or a retired service.
Minutes 16 to 25: Check whether AI crawlers can reach the site
Open yourdomain.com/robots.txt in a browser and look for rules that block the crawlers by name, or a blanket block under User-agent: * with Disallow: /. Then compare it with what each vendor documents.
Read the table by purpose. OAI-SearchBot, PerplexityBot and Googlebot are the three to check first, because a block on any of them can keep you out of ChatGPT search, Perplexity results or Google's AI Overviews and AI Mode. GPTBot concerns OpenAI's model training, and Google-Extended concerns Gemini training and grounding, not Google Search. ChatGPT-User and Perplexity-User visit a page when a person asks a question, and robots.txt may not apply to them.
| Crawler or token | What it controls | What the vendor says about access |
|---|---|---|
| OAI-SearchBot (OpenAI) | Surfaces sites in ChatGPT search. | To appear in ChatGPT search answers, OAI-SearchBot must be allowed. OpenAI's help center adds that your host or CDN must allow traffic from OpenAI's published searchbot IP addresses. OpenAI says it can take about 24 hours after a robots.txt change for its systems to adjust. |
| GPTBot (OpenAI) | Model training. | Disallowing GPTBot tells OpenAI not to use your content for training. OpenAI says each setting is independent. |
| ChatGPT-User (OpenAI) | May visit a page when a person asks ChatGPT a question. | Because these visits are started by a user, robots.txt rules may not apply. It is not used to decide whether content may appear in search, so it is not the agent to check for search visibility. |
| PerplexityBot (Perplexity) | Designed to surface and link websites in Perplexity results. | Perplexity recommends allowing it in robots.txt. It also asks sites to permit its published IP ranges, which matters if a firewall sits in front of your site. |
| Perplexity-User (Perplexity) | A user-triggered fetcher. | It generally ignores robots.txt. |
| Googlebot (Google) | Crawling for Google Search. AI Overviews and AI Mode follow the same rules as Search. | A page must be indexed and eligible to show in Search with a snippet. robots.txt rules for Googlebot are the control for crawling access, and nosnippet, data-nosnippet, max-snippet and noindex limit what is shown. |
| Google-Extended (Google) | Manages whether Google may use your content to train future Gemini models and for grounding in Gemini Apps and Vertex AI. | Google states it does not affect inclusion or ranking in Google Search. |
A clean robots.txt does not prove access, because a CDN or security plugin can still block requests. Run your most important page through the URL Inspection tool in Search Console, which Google recommends for checking the HTML Googlebot received when it crawled the page. If you do not own that access, ask whoever manages your site to check the firewall and the page's robots meta tag.
Your robots.txt review
0 of 6 done
Minutes 26 to 30: Read the result
Sort your spreadsheet by what happened, then match the pattern to the likely cause. Treat it as a snapshot of thirty answers, not a ranking.
- Absent everywhere, even on brand prompts: start with access and indexing.
- Named on brand prompts but absent on category and problem prompts: the engines know you exist but have no reason to pick you. Check whether your pages state plain answers and whether anyone else mentions you.
- Named but with wrong facts: the detail is probably repeated from an old page or directory. Correct the source, then re-run.
- Cited but with no visits: a citation inside an answer is not a click, and the reader may never click at all. The measurement post linked above covers how to measure visibility that does not produce a visit.
- The same competitors every time: open the pages cited for them to see what a quotable answer looks like in your market.
What to do this week
Fix access first, because it is the only cause you can remove in a day. Then pick one answer to improve and re-run the same prompts later.
Five steps for this week
0 of 5 done