How can I see if ChatGPT recommends my company? Manual checks and tools
How can I see if ChatGPT recommends my company? A manual test you can run today, plus the tools that automate it, ranked with sourced pricing.
You can see whether ChatGPT recommends your company in about twenty minutes without buying anything. The catch is that one check proves very little. ChatGPT answers vary by wording, by who is asking, by location and by day, so a single "yes, it named us" or "no, it did not" is an anecdote. Below is a manual method that gets you a useful first read, then the tools that turn it into a repeatable measurement.
The manual test
1. Write prompts a buyer would type, not your brand name. Think "best accounting software for small agencies" or "who can help us with commercial roofing in Leeds." Write at least ten across three groups: category questions, comparison questions that name a competitor, and direct questions about your company.
2. Run each one in a fresh chat with search on. ChatGPT with search browses the web and cites sources. Without it, the answer comes from model memory, which can be out of date. Test both if you can, but note which mode you used.
3. Record what happens. For each prompt, note whether you are named, whether you are recommended or just listed, what position you appear in, which competitors appear, and which sources are cited. A spreadsheet is enough.
4. Check the cited sources. If ChatGPT cites a review site, a Reddit thread or a competitor page and not yours, that tells you where the recommendation is actually being decided.
5. Repeat on another day. Run the same list again a few days later. If answers change a lot, you have learned something about stability.
6. Check crawler access. If your robots.txt or firewall blocks OAI-SearchBot, ChatGPT search cannot fetch your pages. GPTBot and ChatGPT-User are separate bots with separate purposes, so check which ones your rules affect.
Why manual checking stops working
It does not scale. Ten prompts across three countries and two personas is sixty runs, and you need daily repeats to see a trend. It also cannot show you crawler activity or AI referrals in your analytics. That is where tools help.
Tools that automate it, ranked
1. Promptwatch
Promptwatch runs your prompt panel daily and reports Visibility Score, Share of Voice, Sentiment and position in the answer, per model, country and language. Citation analytics shows which pages and domains ChatGPT cites. Agent Analytics shows IP-verified crawler logs, including GPTBot, OAI-SearchBot and ChatGPT-User, with errors per page. Visitor analytics measures ChatGPT-referred visits and conversions. The free Explore plan covers 10 prompts on ChatGPT only, which fits this exact question. Essential is $95 a month, Professional $245 (where crawler logs start), Business $579. A free GEO Inspector Chrome extension shows ChatGPT fan-outs and fetched versus cited sources, and a no-login AI Visibility Report gives a quick scan.
2. AthenaHQ
AthenaHQ has a free Essential tier with 300 credits and a Starter plan at $295 a month with 3,600 credits across 11 models. It adds a GEO score and an Action Center for follow-up.
3. Scrunch AI
Scrunch AI Core is $250 a month for 125 prompts across ChatGPT, Perplexity, Google AI Overviews and Copilot, with site audits and an Agent Experience Platform.
4. Profound
Profound offers a free 7-day trial (50 prompts daily across ChatGPT, Gemini and Google AI Overviews) and custom Enterprise pricing, with an earlier estimate near $40k a year.
5. Add-ons and lighter trackers
Ahrefs Brand Radar includes Custom Prompts on paid Ahrefs plans. HubSpot AEO is $50 a month. Otterly.AI starts at $29 a month for 15 prompts, and Peec AI covers prompt-level monitoring.
Comparison table
| Tool | Free option | Paid from | Shows cited sources | Crawler logs |
|---|---|---|---|---|
| Promptwatch | Explore, 10 prompts, ChatGPT only | $95/mo | Yes | From Professional |
| AthenaHQ | 300 credits | $295/mo | Yes | Not published |
| Scrunch AI | None listed | $250/mo | Yes | Bot traffic reporting |
| Profound | 7-day trial | Custom | Yes | Agent Analytics |
Who should pick which
- You want the quick answer today: the manual test, then a free Explore account to repeat it daily.
- You suspect ChatGPT cannot reach your site: Promptwatch, for the crawler logs.
- You want an action list: AthenaHQ or Promptwatch's Action Board.
What "recommended" can and cannot mean
Being listed is not being recommended. Look at the wording: does ChatGPT say you are a good fit for a stated need, or only mention you in passing? Also, no tool can make ChatGPT recommend you. They show where you stand and what to fix, such as missing pages, blocked crawlers or absent third-party mentions.
A short routine once you have a baseline
Keep the prompt list fixed for a month. Review answers weekly, and read at least a handful of full responses each time, because a score can improve while the wording gets worse. When a gap shows up, pair it with one concrete change: a page that answers the missing question, a crawler rule that was too strict, or a profile on a site ChatGPT keeps citing. Then re-run the same prompts and see whether anything moved. That loop, repeated, is the whole discipline.