Why Perplexity skips you — and how to check without a dashboard

Perplexity is not a reskin of ChatGPT. When a buyer asks who to hire, it retrieves, names, and cites from its own stack. You can be the sentence on ChatGPT and still be absent on Perplexity. Treating “AI search” as one channel hides that split. The check is the same prompt on all three engines — then you fix what that transcript shows, not a generic GEO checklist.

The check (five minutes)

Ask Perplexity the query a customer would type:

• Local: best [category] in [city]
• SaaS / digital: best [category] (no city if you sell worldwide)

Then ask ChatGPT and Google AI the same thing.

Record, per engine:

• Were you named?
• Who was named instead?
• Which URLs did it cite? Your site, a directory, a review platform, a competitor?

If Perplexity cites G2, a round-up, or a competitor’s docs and never your domain, you have a source problem. Writing another homepage paragraph will not put you in a citation it already picked from somewhere else.

What Perplexity is doing that Google is not

Google AI still starts from Google’s index, then picks a passage. Perplexity is built to show sources. That makes the citation list the product: if you are not in it, you were not retrieved as a trustworthy answer for that query.

PerplexityBot in robots.txt is the floor. Block it and you are arguing with a system that never fetched you. Allow it and you still have to be a clear entity with an extractable answer and, often, a third-party page it already trusts.

What to do this week

1. Confirm PerplexityBot is not disallowed at https://yoursite.com/robots.txt. Use the free crawler check if you do not want to read the file by hand.
2. Put a 75–150 word answer at the top of the page that should win the query.
3. Re-run the three-engine prompt. Keep the Perplexity citation list. Those domains are the directories and review sites worth being on — not a generic “get listed everywhere” list.
4. If a named source already has your product and Perplexity still cites a competitor there, the job is connecting that listing (exact name, website, sameAs), not collecting more reviews for sport.

If you want that transcript done across ChatGPT, Google AI, and Perplexity — who got named, which URLs were cited, 12 steps over 30 days — that is the one-time diagnostic this site sells. One payment. About five minutes after intake.

Frequently asked questions

If ChatGPT names me, why would Perplexity skip me?

They do not share one index. ChatGPT leans on Bing and its own browsing. Google AI uses Google’s index. Perplexity retrieves and cites from its own stack. Passing one engine is not a pass on the others.

What should I ask Perplexity?

Use the same buyer prompt you use elsewhere: best [category] in [city], or best [category] with no city if you sell worldwide. Write down whether you were named, who was named instead, and which URLs it cited.

Does PerplexityBot in robots.txt fix this?

If PerplexityBot is blocked, stop and unblock it. If it is allowed and you are still missing, the gap is entity, citable answers, or third-party sources Perplexity already trusts — not another robots tweak.

Should I write a Perplexity-only page?

No. Lead with a 75–150 word answer on the money page, keep crawlers allowed, and get listed on the sources the live answers already cite. That helps every engine, including Perplexity.