Skip to content

AI visibility · How-to

Perplexity for business: how practices appear in Perplexity answers

How Perplexity finds, cites and recommends local practices, what its crawler needs from your site, and how to test answers for your specialty and city.

By Medfy Editorial TeamUpdated Facts verified 7 min read

Perplexity for business starts with one useful fact: Perplexity always shows its sources. Every answer comes with numbered links, so a practice can see exactly which pages shaped what a patient was told. That makes Perplexity the easiest AI assistant to audit. This guide covers how it answers, how to let its crawler in, how to test it for your specialty and city, and how Gemini and Claude compare.

Search demand is small and specialized: "perplexity seo" gets 110 searches a month and "perplexity for business" 70, with a DataForSEO keyword difficulty of 55 (Google Ads Keyword Planner data via DataForSEO, US, 12-month average to Sep 2026). The people searching are usually trying to answer the question this page answers.

How Perplexity answers questions

Perplexity describes its process in its own help center (Perplexity (opens in a new tab)):

  • It searches the web in real time for each question rather than relying only on what a model learned in training.
  • It gathers information from authoritative sources such as articles, websites and journals, and condenses it into one answer.
  • "Each answer includes numbered citations linking to the original sources."
  • It runs on several language models, and by default picks the one it judges best for the question. Pro subscribers can choose a model.
  • Its Research mode "performs dozens of searches automatically, reads hundreds of sources" and writes a longer report.
  • It keeps context within a conversation, so a patient's follow-up ("which of those take Medicaid?") builds on the first answer.

Two consequences for a practice: answers reflect what's on the web this week, and the model behind them can differ from one question to the next. Both make answers vary.

If you'd like a starting point before testing by hand, the free marketing audit checks whether PerplexityBot and other AI crawlers can read your site.

Why Perplexity's visible sources help practices

With most assistants you see the conclusion. With Perplexity you see the evidence, which turns a vague worry ("AI doesn't know about us") into a fixable list:

  • You're named and your site is cited: your pages are doing their job.
  • You're named but a directory is cited: the directory is carrying your reputation. Check it's accurate, and strengthen your own page.
  • A competitor is named and their site is cited: open that page. It usually answers the patient's exact question more directly than yours.
  • No practice's own site is cited: the answer is built from directories and reviews. Your listings matter more than your blog here.

Perplexity is also where the original research on generative engine optimization (opens in a new tab) was field-tested. In that 2023 study, rewriting source pages to cite sources, add quotations and add statistics improved visibility on Perplexity.ai by up to 37%, while keyword stuffing did 10% worse than no change.

Let PerplexityBot read your site

Perplexity names two agents in its crawler documentation (Perplexity (opens in a new tab)):

  • PerplexityBot is "designed to surface and link websites in search results on Perplexity." Perplexity says it "is not used to crawl content for AI foundation models."
  • Perplexity-User fetches a page when a user's question calls for it. Perplexity says this fetcher "generally ignores robots.txt rules" because a person requested it.

To make sure you can be cited:

  1. Open yoursite.com/robots.txt and check for rules blocking PerplexityBot or all bots.
  2. Allow PerplexityBot. Perplexity recommends allowing it so your pages can appear in its results.
  3. Check your host or CDN. If it blocks unknown bots, allow requests from Perplexity's published IP list.
  4. Keep key facts in page text. Hours shown only in an image, or insurance listed only in a PDF, are easy to miss.
User-agent: PerplexityBot
Allow: /

Our guide to AI crawlers and robots.txt covers every assistant's bots in one place.

The sources Perplexity tends to cite for local care

Perplexity doesn't publish how it chooses sources for local questions, so don't trust anyone who claims to know its formula. What you can do is observe. When you test, sort every cited link into one of these buckets:

Source typeWhat to checkWhat to do
Your own websiteWhich page is cited, and is it the right one?Strengthen that page; add a page if the service has none
Directory and listing pagesName, address, phone, hours and insuranceCorrect errors; claim unclaimed profiles
Review platformsRating, volume, recencyAsk for reviews steadily on the platforms cited
Hospital and health-system pagesYour provider bios and affiliationsAsk the system to update bios and link your site
Local news and community pagesWhether they mention you at allShare genuine news: a new service, a new clinician
Forums and Q&A threadsWhat patients say about local optionsDon't post as a fake patient; improve the service they discuss

After four weeks, the bucket counts tell you where to spend time. If 70% of citations for "best physical therapist in Spokane" are directories, your listings are the priority. That figure is illustrative, not a benchmark.

Testing Perplexity for your specialty and city

  1. Pick six to ten fixed prompts a patient would type: "best physical therapist in Spokane", "physical therapy Spokane that takes Premera", "sports physical therapist near Gonzaga".
  2. Use the default settings, logged out where possible, so you see what a typical patient sees.
  3. Write down where you're testing from. Location and phrasing both change local answers.
  4. Record four things: practices named, their order, every cited URL with its bucket, and any wrong facts about you.
  5. Ask one follow-up a patient would ask, such as "which of these are open Saturdays?", and note whether you stay in the answer.
  6. Repeat weekly on the same day and compare a month of results.

Alongside the tests, watch for visits referred from perplexity.ai in your analytics. We found no public survey that breaks down patient use by assistant; KFF's March 2026 poll (opens in a new tab) found 32% of US adults used AI chatbots of any kind for health information in the past year.

Illustrative example: A fictional practice, Riverside Physical Therapy in Spokane, appears in two of four weekly answers for "best physical therapist in Spokane". Every citation for it is a directory page listing an old suite number. After fixing the directory and adding a page on sports injury rehab, the practice records which source Perplexity cites in the following weeks.

Common mistakes

  • Testing once and treating the answer as permanent.
  • Blocking PerplexityBot while trying to block AI training. PerplexityBot isn't a training crawler, according to Perplexity.
  • Ignoring the citations. The links tell you what to fix; the answer text alone doesn't.
  • Posting fake reviews or forum comments. They mislead patients and can breach platform rules.

Gemini and Claude: what's similar, what's different

The same habits carry over, but the plumbing differs. Each provider's own documentation, as of this page's verification date:

ItemPerplexityGemini appClaude
Web accessReal-time web search for answers"Some Gemini responses are grounded on Search results" (Google (opens in a new tab))Web search tool; on Team and Enterprise plans an owner must turn it on (Anthropic (opens in a new tab))
CitationsNumbered citations on every answerNot described on the privacy page checked"Every response includes citations" with web search
LocationNot documentedGeneral area from IP address, or Home and Work addresses; precise location with permissionMay use location inferred from your IP address for local results
Crawlers and controlsPerplexityBot (search), Perplexity-UserGooglebot; the Google-Extended token covers Gemini training and grounding but not Google Search (Google (opens in a new tab))Claude-SearchBot (search), Claude-User (user requests), ClaudeBot (training) (Anthropic (opens in a new tab))

Two practical notes. Anthropic says blocking Claude-SearchBot "may reduce your site's visibility and accuracy in user search results", so block ClaudeBot if you want to opt out of training, not the search bot. And because Google says Google-Extended covers grounding in Gemini Apps as well as training, blocking it may keep your pages out of some Gemini answers, even though it doesn't affect Google Search.

For ChatGPT, which documents its crawlers differently, see our guide to getting recommended by ChatGPT. Why some pages get cited and others don't is covered in AI citations, and the wider picture in AI visibility for practices.

Where Medfy fits

Medfy's AI visibility tracking runs weekly checks of Perplexity, ChatGPT, Gemini, Claude and Google AI Overviews using the questions patients ask, and shows whether you're named, who is recommended instead and how that changes. If you're comparing options, our guide to AI search tracking tools explains what to look for. To check where you stand now, start with the free marketing audit.

Frequently asked questions

Does Perplexity recommend local businesses?

It can. Perplexity answers questions by searching the web in real time and citing its sources, so when a patient asks for a practice in a city, it can name the practices those sources describe. Perplexity's help center doesn't explain how it ranks local options, so test your own specialty and city weekly and record which practices and sources appear.

How do I get my practice cited in Perplexity?

Let PerplexityBot crawl your site, then give it something worth citing: a page per service that answers patient questions directly, with clinical facts linked to MedlinePlus or PubMed and local facts such as insurance and hours in plain text. Keep your listings consistent, because directories are often cited too. Nothing can make a citation certain; answers vary by phrasing, location and day.

Should I let PerplexityBot crawl my site?

For most practices, yes. Perplexity says PerplexityBot exists to surface and link websites in its search results and is not used to crawl content for AI foundation models. Blocking it removes a way for patients to reach your site from Perplexity answers. If your concern is AI training, block the training crawlers other companies name instead.

Do patients use Perplexity to find doctors?

Some likely do, but we have found no public data on how many patients use Perplexity specifically. KFF's March 2026 poll found 32% of US adults used AI chatbots for health information in the past year, without breaking that down by product. The best evidence is your own analytics: look for visits referred from perplexity.ai and watch the trend.

Sources

  1. How does Perplexity work? (opens in a new tab) — Perplexity Help Center. Accessed October 22, 2026.
  2. Perplexity Crawlers (opens in a new tab) — Perplexity. Accessed October 22, 2026.
  3. GEO: Generative Engine Optimization (Aggarwal et al., arXiv 2311.09735, accepted to KDD 2024) (opens in a new tab) — arXiv. Accessed October 22, 2026.
  4. Gemini Apps Privacy Hub (opens in a new tab) — Gemini Apps Help. Accessed October 22, 2026.
  5. Google's common crawlers (Google-Extended) (opens in a new tab) — Google for Developers. Accessed October 22, 2026.
  6. Does Anthropic crawl data from the web, and how can site owners block the crawler? (opens in a new tab) — Anthropic (Claude Help Center). Accessed October 22, 2026.
  7. Enable and use web search (opens in a new tab) — Anthropic (Claude Help Center). Accessed October 22, 2026.
  8. Poll: 1 in 3 Adults Are Turning to AI Chatbots for Health Information, Equaling the Share Who Use Social Media for Health (opens in a new tab) — KFF. Accessed October 22, 2026.
All pages

Free AI marketing audit

See how your practice shows up — free

Scores for your website, SEO, Google listing, AI visibility, social and ads, with your top five fixes. It reads public information only.