Key takeaways
- Perplexity searches the web for every answer: 17.4 source links per answer in our data, against 5.2 for ChatGPT.
- Let PerplexityBot crawl your pages; a robots.txt block keeps you out of its sources.
- One page per buyer question, the answer in the first two sentences, backed by dated facts and a table.
- In 50.7% of Perplexity answers naming a brand, the brand's own site was not cited: work the third-party pages too.
To optimize content for Perplexity, let PerplexityBot crawl your pages, answer one buyer question per page in the first two sentences, back the answer with dated facts and sources, and get mentioned on the third-party pages Perplexity already cites. Perplexity searches the web for every answer, so a page you publish or fix today can be picked up without waiting for a model update.
In Howseen's data, 100% of 1,098 Perplexity answers had at least one source, with 17.4 source links per answer on average, against 5.2 for ChatGPT (our study of 5,331 AI answers). More links per answer means more room for a page that answers the question well.
How does Perplexity choose the pages it cites?
Perplexity runs a web search for the question, reads the pages it retrieves and writes an answer with numbered sources. Three facts from our data shape what to do:
- It spreads citations widely. Its 10 most cited domains take 8.7% of its links, against 18.0% for ChatGPT, across 3,190 distinct domains: smaller sites get a share.
- It names brands through other people's pages. In 50.7% of the Perplexity answers that named a tracked brand, no page of that brand's own site was among the sources (fan-out study).
- Forums and video are a small part of it. Perplexity cited Reddit in 4.9% of its answers and YouTube in 0.5% in the same data.
What are the steps to optimize a page for Perplexity?
- 1. Let PerplexityBot in. Perplexity writes that PerplexityBot "is designed to surface and link websites in search results on Perplexity" (Perplexity crawler documentation). A robots.txt block keeps your pages out. Test it with the AI crawler checker.
- 2. One page per buyer question. Put the question in the title or an H2, and the answer in the first two sentences, with the key fact in it.
- 3. Add the facts that back the answer. Prices, limits and numbers with their source and date, and a table when buyers compare options. The Princeton GEO study reports that such methods "can boost visibility by up to 40% in generative engine responses" (GEO, arXiv).
- 4. Serve the text as HTML. Content that only appears after JavaScript runs may not be read by a crawler.
- 5. Get on the pages it already cites. List the domains Perplexity cites on your questions, then earn a mention or a listing there: reviews, comparisons, directories.
- 6. Date the page and keep it current. Show the update date, and change the facts when your prices or features change.
What does a page built for Perplexity contain?
| Element | What to write | Why it helps |
|---|---|---|
| First two sentences | The direct answer, with the key fact | A reader or an engine gets the answer without scrolling |
| H2 headings | The variants of the question buyers type | Covers the several searches behind one question |
| Table | Prices, features or limits side by side | Facts that are easy to extract and compare |
| Sources | Links to official pages, with the date checked | The answer can be verified |
| FAQ | 4 to 6 short questions and answers | More phrasings of the same need |
| Update date | A real date, changed when facts change | Shows the facts are current |
My take, from building Howseen: Perplexity is the engine where our own brand shows up most, 5 of our 50 tracked questions in the November 14, 2026 scan. It reads new pages fast, which makes it the first place to check whether a page change worked.
How do you know if your Perplexity optimization worked?
Run the same buyer questions in Perplexity on a schedule and keep every answer with its sources, so you see the day your page enters the list. By hand, that is a private window and a spreadsheet. Howseen does it on Pro (€79 per month) every 3 days and on Business every day: see how to track Perplexity citations, and read the same logic for AI share of voice.
How long does Perplexity take to cite a new page?
Perplexity publishes no delay. Because it searches the web for each answer, check your questions again a few days after the page is crawled and indexed, and keep checking: answers change between runs.
Does blocking PerplexityBot remove my site from Perplexity?
It can keep your pages out of the sources. Perplexity describes PerplexityBot as the crawler that surfaces and links websites in its results, so a robots.txt block works against being cited.
Does Perplexity-User respect robots.txt?
Perplexity says Perplexity-User supports user actions, such as fetching a page a user asked about, and that because a user requested the fetch it generally ignores robots.txt rules.
Do I need an llms.txt file for Perplexity?
Perplexity's crawler documentation does not mention llms.txt. Crawl access, a clear answer at the top of the page and mentions on the sites it cites matter more.
How do I track my Perplexity citations?
Run your buyer questions in Perplexity on a schedule and store every answer with its sources. Howseen does it every 3 days on Pro and every day on Business, and flags the questions where your site is cited.