How to show up in Perplexity answers (2026)

Perplexity is the most transparent of the major answer engines. It shows its sources alongside the answer, so when you are left out, you can usually see who was used instead. That makes it a good engine to start with when a client wants to understand why AI answers skip them.

Showing up in Perplexity can mean two things. Your brand can be named in the answer text, or one of your pages can appear as a source. Our visibility score counts the first, because a buyer reading the answer sees names before they look at sources. We track the second separately, since it is often what produces the first.

Here is the sequence we follow.

Step 1: let PerplexityBot in

Perplexity documents its crawlers on its crawler page. It describes PerplexityBot as the crawler that surfaces and links websites in Perplexity's search results, and says it is not used to crawl content for AI foundation models. Perplexity recommends allowing PerplexityBot in robots.txt and permitting requests from its published IP ranges if you want to appear in results.

That separation is useful when a legal or security team wants to block AI training. Blocking PerplexityBot does not opt you out of training, because Perplexity says that is not what it does. It mostly opts you out of being linked.

Open your robots.txt and look for any rule that names PerplexityBot, and any wildcard rule that would catch it. Perplexity says changes can take up to 24 hours to be reflected, so do not retest the same afternoon.

Step 2: check your firewall and CDN

Robots.txt is only part of access. Bot protection can challenge or block a crawler your robots.txt allows. Perplexity's documentation addresses this directly, with configuration steps for Cloudflare and AWS WAF. The approach it describes combines two conditions: the user agent contains PerplexityBot or Perplexity-User, and the source IP is in the ranges Perplexity publishes as JSON files. The action is Allow, set at a higher priority than your blocking rules.

Perplexity warns that those IP lists are updated regularly and suggests automating the refresh. A hand-copied list of IP ranges will go stale, and the failure looks exactly like Perplexity ignoring your content.

There is a second agent to know about. Perplexity-User fetches pages when a user's question calls for it. Perplexity says it is not used for crawling, and because a user requested the fetch, it generally ignores robots.txt. If a page must stay private, put it behind authentication. A Disallow line will not keep it out of answers.

Step 3: confirm in the logs

Once the rules are in place, check your server or CDN logs. You want PerplexityBot requesting the pages that matter, such as product, pricing, comparison and use-case pages, and getting a 200. Perplexity notes that firewall changes take time to propagate and recommends watching the logs, which is good advice for any rule change.

If key pages get no requests at all, the issue is discovery: weak internal links or missing sitemap entries. If they get requests and errors, it is access. Different problems, different fixes.

Step 4: read the sources on your missing prompts

Now the useful part. For each tracked prompt where Perplexity does not name you, open the answers and list the sources. Sort them roughly:

• Competitor pages
• Third-party roundups and review sites
• Community threads and videos
• Your own pages

That last group is the interesting one. If Perplexity cites your page and still names a competitor, the page is being used for something, perhaps a definition or a statistic, without making the case for your product. Check whether the cited passage mentions your product at all. Often the fix is on that page rather than a new one.

Step 5: publish pages worth citing

For prompts where no page of yours is cited, look at what was. Cited pages tend to answer the question in their opening lines, in the buyer's terms, with specifics that can be lifted into an answer. Comparison pages, honest use-case pages and clear pricing explanations fill gaps that brochure-style product pages leave.

Tie each new or revised page to the prompts it is meant to win. Without that link, you cannot tell later whether it worked.

Perplexity's crawler documentation does not mention llms.txt or any special markup, so we do not sell either as a Perplexity tactic. Clean HTML with the answer in visible text is what matters.

Step 6: work the third-party sources

Some prompts will be answered from the same few outside domains no matter what your site says. If one roundup or one community thread keeps appearing across a group of your prompts, that is where the effort goes: accurate listings, genuine coverage, and open participation as a named company representative where the community allows it. We do not buy placements or post undisclosed promotion. It breaks most community rules and it damages the source that made the citation worth having.

This is the slowest step, so start it early.

Step 7: measure Perplexity on its own

Score Perplexity separately from the other engines, for at least two reporting periods, before concluding anything. Perplexity reports 22M+ monthly active users according to Promptwatch's published figures. That is a reason to track it, not a traffic forecast, so measure the visitors it actually sends you.

How we track Perplexity in Promptwatch

We measure Perplexity, like every engine, on Promptwatch. Prompt tracking monitors Perplexity's real interface, so the answers we score match what a buyer sees. Every paid plan runs all AI models, Perplexity included. The free Explore plan is ChatGPT only, so Perplexity tracking starts from Essential at $95 a month.

Citation analytics shows the exact pages and domains Perplexity cites for each prompt, including Reddit and YouTube citations, which is step 4 without the copy and paste. Agent Analytics logs PerplexityBot requests in real time, flags errors, and shows the crawl-to-citation path for a page, which covers steps 1 to 3. On brand plans, crawler logs are listed from Professional ($245 a month). Visitor analytics, added with a small script or a GTM template, shows the visits Perplexity sends and whether they convert.