Perplexity SEO: How to Get Your Brand Cited and Recommended

Perplexity answers a buying question with a short list of options and a row of numbered sources. Your brand can win in two places there: named in the answer, or linked as one of the sources. Perplexity SEO is the work of earning both, in a fixed order: make sure Perplexity can reach your pages, give it something worth quoting, then measure whether it does.

This is the step-by-step guide. If you want the research on what Perplexity tends to cite, such as its heavy use of comparison pages and niche directories, read how Perplexity chooses which brands to cite alongside it.

Mentions and citations are separate goals

Classic SEO asks where your page ranks. On Perplexity the better questions are whether the answer names you and whether it links to you, because the two do not always travel together. A review site can be cited in an answer that recommends your rival; your brand can be recommended on the strength of other people's pages without your site appearing in the source list. Track them as two numbers from the start.

Everything about crawlers below comes from Perplexity's own documentation, checked in November 2026. The content and measurement advice is our editorial recommendation, to be tested against your own data.

Step 1: let Perplexity's crawlers in

Perplexity's crawler documentation describes two user agents, and they behave differently.

• PerplexityBot is the crawler behind Perplexity's search results, where it surfaces and links websites. Perplexity states that it does not collect pages for foundation-model training, and advises site owners to allow it in robots.txt.
• Perplexity-User visits a page when a person's question needs it. Because the person asked, Perplexity says this fetcher generally ignores robots.txt.

So there are two checks to make. First, read your robots.txt for any rule that stops PerplexityBot, including a broad User-agent: * block you may have inherited from a template. Second, check your firewall or CDN. A bot-blocking preset can refuse both agents whatever robots.txt says. Perplexity publishes separate IP ranges for each agent and setup steps for Cloudflare and AWS WAF, and it advises matching each user agent with its own IP list rather than one combined list. It also notes that Cloudflare's Bot Fight Mode cannot be skipped with a custom rule, so it may still challenge the bots.

Allow up to 24 hours for robots.txt changes to register, which is the window Perplexity gives. For a quick reading of where you stand, the free AI crawler checker tests your robots.txt against PerplexityBot and other AI crawlers.

Two limits are worth stating plainly. Access makes citation possible; it does not guarantee it. And blocking AI crawlers can be a deliberate, legitimate policy, with trade-offs that should you block GPTBot? sets out vendor by vendor. This guide assumes you want to be cited.

Step 2: publish the answers buyers are asking for

Perplexity retrieves pages live and lifts passages from them, so the page that wins is usually the one that answers the question outright. These are recommendations to test, not documented ranking rules.

Map the questions first. List the questions that decide a sale in your category: price, how you compare with the two names you meet most often, reliability, after-sales terms, compatibility. Then find the page on your site that answers each one. Gaps on that list are your content plan.

Lead with the answer. Give each page a heading phrased the way buyers ask, and answer in the opening sentence or two. Follow with the evidence: numbers, dates, named models and conditions.

Keep facts in HTML text. Terms buried in a scanned brochure or a slide image are harder for any reader to lift. Put prices, specifications and policy terms on the page itself, and date anything that changes.

Say the uncomfortable parts yourself. A page that admits who your product is not for, or what costs extra, gives Perplexity a balanced source to quote at the objection stage instead of a forum complaint.

Fix contradictions. If your pricing page, marketplace listings and a two-year-old press release quote different prices, any of them may end up in an answer. Retire or update the stale ones.

An invented example makes the point. Two fictional mattress makers, Example Sleep and Sample Beds, are similar on price. A buyer asks Perplexity which has the better trial period. Sample Beds has a returns page headed "How long is the trial?" with the answer in its first line. Example Sleep's trial terms sit on page 14 of a downloadable care guide. Nothing here predicts what Perplexity will say; the point is that one brand has made the answer easy to find and the other has not, and that is the part a team controls.

Step 3: widen the set of pages that mention you

Perplexity assembles answers from many sites, and much of what it says about a brand comes from pages the brand does not own. Find the review sites, comparison articles, trade directories and community threads that appear in Perplexity's source lists for your category's questions, and check what they say about you. Where a listing is incomplete or out of date, correct it at the source; Perplexity can only repeat what it finds.

Step 4: measure citations and the final pick

Set up a routine you can repeat. Choose a fixed set of buyer prompts, run them in Perplexity on a schedule, and for each answer record three things: were you named, was your site among the sources, and were you the pick when the buyer asked which to choose. A single run tells you little, because answers vary from one asking to the next; look for changes that hold across several weeks. Track two or three named competitors on the same prompts so you can tell your own movement from a reshuffle across the category. The prompt tracking entry in our glossary describes the approach in brief.

Pay most attention to the follow-ups. If you are named in the first answer and gone when the buyer asks about reliability or value, open those answers, note which sources Perplexity used instead of yours, and check whether the fact the buyer asked about is published clearly anywhere.

Where Vouch fits

Perplexity is one of the seven assistants Vouch tracks every day with single prompts, alongside ChatGPT, Claude, Gemini, Copilot and both of Google's AI features, AI Overviews and AI Mode. It is also one of the four where Vouch runs Threads, its six-question buyer conversations, three times a week, to find the question where your brand is dropped and the rival named in its place. How the questions are put to each assistant is set out in the methodology.

Pricing runs from $39 a month on the smallest plan to $599 on the largest. New accounts are by invitation for now, with a 14-day free trial to start; put your name on the waitlist.