Most advice on ranking in Perplexity reads like a Google checklist with a new name. Some of it holds up. A lot of it rests on one vendor's study, or on a figure that does not come from Perplexity at all.
This page separates what Perplexity documents, what third-party studies suggest, and what nobody has shown. Then it gives you a method you can run in an afternoon: check access, rewrite for quotability, earn mentions, and test the same prompts every week. Our wider guide is the AI visibility hub, and the service view sits at generative engine optimization services.
01How Perplexity picks sources
Perplexity answers a question by searching the web, fetching pages and citing the ones it uses in the response. Its own crawler documentation describes two bots: PerplexityBot, which surfaces and links sites in Perplexity search results, and Perplexity-User, which visits a page when a person's question calls for it and may link that page in the answer. Neither document explains how results are ranked.
That gap matters. Every ranking-factor list you see, including the percentages, comes from third-party analysis and is not a published rule. Wix, in its Perplexity guide, reports that about 60% of Perplexity citations overlap with Google's top 10 for the same query, citing BrightEdge. Use that as a sign that ranking well in Google helps, not as a formula.
| Claim | Status | Source |
|---|---|---|
| PerplexityBot surfaces and links sites in Perplexity results | Documented by Perplexity | Perplexity crawler guide |
| Perplexity-User fetches pages for a person's question and generally ignores robots.txt | Documented by Perplexity | Perplexity crawler guide |
| About 60% of citations overlap with Google's top 10 | Reported by Wix, citing BrightEdge | Wix Studio, How to Rank in Perplexity |
| Reddit is 46.7% of Perplexity's top 10 citations, against 11.3% for ChatGPT | Reported by Wix, citing Profound data | Wix Studio, How to Rank in Perplexity |
| Ranking weights for freshness, authority or schema | Not published | No primary source found |
02Check that Perplexity can read you first
Before changing any content, confirm the bots are allowed in. Perplexity recommends allowing PerplexityBot in robots.txt along with its published IP ranges so your site can appear in results. It also says changes to robots.txt can take up to 24 hours to take effect, and that each robots.txt setting works independently.
Two traps catch real sites. First, a firewall or bot-protection rule can block a bot that your robots.txt allows, which is why Perplexity publishes setup guidance for Cloudflare and AWS and recommends matching both the user agent and the IP ranges. Second, Perplexity says Perplexity-User generally ignores robots.txt rules because a person requested the fetch, so blocking by robots.txt will not stop those visits.
Five-minute access check
0 of 5 doneIf you want to compare the same access issue for ChatGPT, our guide to fixing wrong ChatGPT information has a table of OpenAI's bots.
03Write each page so one passage can be quoted
MyMentions advises making key passages self-contained so they can be quoted on their own, and we agree: every section needs a first sentence that answers its own heading. Put the answer in the first one or two sentences under a heading phrased as the question a person would type, then support it with a table, steps or a short list. A passage that only makes sense with the paragraph before it cannot be quoted on its own.
Length is not the lever. Ahrefs, studying Google AI Overviews, found 53.4% of cited pages were under 1,000 words and that word count barely correlated with being cited. That is Google data, not Perplexity data, but it supports the same practice: make sections self-contained instead of making pages longer. Keep structured data only where it matches the visible page, as Google advises for its own features.
04Earn mentions on pages Perplexity already reads
Your own pages are half of it. The other half is whether other pages describe you the same way. Ahrefs found branded web mentions correlate 0.664 with appearing in AI Overviews across 75,000 brands, the strongest factor it measured, against 0.218 for backlinks. That is a Google result, but the principle carries: being named in many places helps a system that checks several sources.
Find where mentions come from by asking Perplexity your category prompts and opening every cited source. The pages that appear for your category are your target list. Wix reports that review sites and 'best of' roundups carry heavy weight in Perplexity recommendations, citing a FirstPageSage survey, and that Reddit appears often in citations, citing Profound. Treat those as leads to test.
List the cited pages
For each of your 10 core prompts, copy every URL Perplexity cites. Group them by type: roundup, review site, forum, news, vendor page.
Count the types
If six of ten prompts cite roundups, roundups are your target. If forums dominate, community answers are.
Pitch or participate honestly
Ask roundup authors for a fair listing with a reason, and answer forum questions with useful detail and no sales pitch. Never buy fake reviews or pose paid placement as editorial.
Keep one description everywhere
Use the same one-line description on every profile you control, so a system reading several pages sees one consistent company.
Recheck weekly
Rerun the same prompts and note whether new sources now cite you.
05Freshness: what is known and what is not
Freshness probably matters for time-sensitive prompts, but the numbers floating around are thin. Nick Lafferty's guide, which links to Profound, reports a Profound analysis of 883,000 pages across seven engines in which the median page lost half its citation share soon after peaking. That is a vendor study of several engines, not of Perplexity alone, and the same article gives other weights with no source.
A defensible rule is to keep pages that carry prices, comparisons and facts correct, and to leave evergreen explanations alone unless they are wrong. Changing a date on a page without changing its content is not a fix, and we do not recommend it.
06A weekly test you can run
Ranking in Perplexity is a test-and-compare job, so run one weekly loop. MyMentions describes the same weekly loop in its playbook: log the prompt, the cited URLs, the position and the competitors for each core prompt, then fix and expand.
Audit
Check bot access, then run 10 to 20 prompts built from sales calls and support tickets. Log which are answered with citations, and which of those name you.
Fix
Rewrite the pages that should be cited so each section opens with its answer. Correct wrong facts at their sources.
Expand
Add mentions on the pages that Perplexity already cites for your category. Add pages for prompts you do not yet answer.
Measure
Rerun the same prompts. Track presence, citation and accuracy as separate numbers, as in our guide to measuring AI visibility.
Scroll sideways to follow the stages.
| Prompt | Cited URLs | Position in answer | You named? | You linked? | Competitors named |
|---|---|---|---|---|---|
| Best scheduling software for a small dental practice | 5 URLs, 2 roundups | Not listed | No | No | Three |
| Family lawyer in Leeds for child arrangements | 4 URLs, 1 directory | Third | Yes | Yes | Two |
07What is unproven
Several popular tactics have no evidence behind them for Perplexity. Perplexity gives no rule that FAQ schema, llms.txt or a specific word count gets you cited; Google says plainly that no special AI file or markup is required for its own AI features, and we treat llms.txt as optional, cheap and unproven. Wix suggests adopting llms.txt, joining Perplexity's Merchant Program (free to join through a form, for product data) and publishing on Perplexity Pages. The Merchant Program is relevant to online shops, see AI SEO for ecommerce. Nothing we opened shows any of these lifts citations for a service business.
Work on Perplexity first if
- Your buyers are researchers who ask detailed questions
- You already rank and want to be cited, not only listed
- You can earn mentions on review and roundup pages
Put it behind other work if
- Your pages are not indexed in Google yet
- Your brand facts are wrong across profiles
- Your buyers do not use AI search tools for your category
What most sites do today
- Copy a Google checklist and call it Perplexity SEO
- Block AI bots by default and never test
- Publish long pages with the answer buried
- Chase Reddit threads with promotion
- Never rerun the same prompts, so nothing is learned
What we build
- An access check for PerplexityBot, firewall rules and page rendering
- Pages whose sections each open with a quotable answer
- A target list of the sources Perplexity cites for your category
- Honest mentions, and one description everywhere
- A weekly prompt log kept alongside our AI visibility audit
