Make sure PerplexityBot can crawl your site, publish clear pages on the questions your buyers ask, and earn a place in the sources Perplexity already cites in your field. Perplexity documents crawler access, but not how it picks sources. So nobody can promise a citation. Fortunately its citations are visible, so you can study them.
What does Perplexity document?
Perplexity documents how answers are shown and how its crawlers behave. It says little about selection. Its help page says it searches the internet for information from authoritative sources such as articles, websites and journals, and that each answer has numbered citations linking to the originals. It doesn't define "authoritative" or explain how sources get picked.
On crawling, the crawler documentation and the robots.txt article are more specific.
| Crawler | Purpose | robots.txt |
|---|---|---|
| PerplexityBot | Surfaces and links websites in Perplexity's search results. Perplexity says it isn't used for training foundation models | Respected. Blocked sites' text isn't indexed, but the domain, headline and a short factual summary may be |
| Perplexity-User | Fetches a page when a user's question calls for it | Generally ignored, because a user requested the fetch |
Perplexity recommends allowing PerplexityBot if you want to appear in results, and publishes its IP ranges. Changes can take up to 24 hours to apply. It also works with third-party crawlers for its index, which it says must respect robots.txt too.
What should you check first?
- robots.txt. Is PerplexityBot allowed, or caught by a wide block on bots?
- Firewall and CDN. Does your security layer let it through? This fails more quietly than robots.txt.
- Readable text. Is the important content on the page as text?
- A control test. Ask Perplexity about something only your site states. Does it find that?
How do you find out what gets cited in your market?
Perplexity shows numbered sources with each answer, so you can look instead of guessing. We'd use that as your main method. It's more relevant to you than any generic list of ranking factors.
- Write ten to twenty buyer questions, as in this method
- Ask each question and record every cited source
- Group them by type: competitors, trade media, directories, forums, reference sites
- For each group, decide whether you can realistically appear there
- Repeat after a month and only trust what returns
That gives you a much shorter to-do list. If trade media dominate, a real contribution to one trade title may be worth more than five blog posts. If competitors' guides are cited, compare their pages with yours on the same question. These are our suggestions from reading the pattern, not rules from Perplexity.
What seems to make a page citable?
We can only offer informed assumptions here. In the GEO paper, which also ran tests on Perplexity, adding citations, quotations and statistics improved visibility, while keyword stuffing did little. That was a research setting in 2023, not a statement from Perplexity. It fits common sense: a page that gives sourced, specific facts is easier to use as evidence. So:
- State the answer directly under the question
- Add facts you can back up, and name your source
- Keep dates and prices current. Stale pages can produce stale answers
- Publish what you know from real work, and don't rewrite what's already out there
Be careful with made-up numbers and manufactured quotes. They might get picked up, and then you've spread an error under your own name.
What if Perplexity gets you wrong?
You can report it with the flag icon below an answer or through a support ticket, including the URL of the query, the error and the expected result. Then fix the source, because that's where a repeat comes from. See what to do when AI gets your business wrong.
