How to rank on Perplexity: get your business cited
Short answer: there is no ranking list to climb on Perplexity and no form to submit your site to. Perplexity answers a question by finding web pages and linking the ones it used next to the answer, so "ranking" means being one of those linked pages. You can't buy that spot. What you can do is let its crawler read your site, put the answer to the buyer's question in the first lines of a page, keep your facts the same everywhere, and get named on the pages it already links to. Nobody outside Perplexity can promise a result. Checked 14 October 2026.
The shared steps are covered in the ChatGPT guide; the steps below focus on Perplexity.
What Perplexity does differently
Perplexity is built around showing where an answer came from. Each answer has links to the pages it drew on, so you can see exactly which sites "won". That makes it the easiest AI tool to learn from: you don't have to guess who it trusts, you can read the list.
Two things follow. First, the citations are your research. Ask the questions your customers ask, then open every link in the answer. Second, because answers are written from pages fetched at the time, a page you fix can start to appear without waiting for a model to be retrained. Perplexity hasn't published how it picks between pages, so we don't claim to know its ranking signals. Everything below is either from Perplexity's own documentation or a method you can test on your own questions.
Step 1: ask Perplexity the questions your customers ask
Open Perplexity in a private window and ask these, swapping the square brackets for your details. Ask each two or three times, because answers change between asks.
What are the best [type of business, e.g. cafés for brunch] in [town]?
I need a [service, e.g. emergency plumber] in [suburb or town]. Who would you recommend, and why?
How much does [job, e.g. a blocked drain repair] cost in [town]?
Is [business name] in [town] any good? What do customers say?
What are the opening hours, address and prices for [business name] in [town]?
The first three don't name you, so they test whether a stranger would be sent to you. The last two name you, so they test what Perplexity knows. Count them separately.
For every answer, write down: did it name you, what did it get wrong, and which pages it linked. Use a simple sheet:
| Question | Date | Named you? | Wrong or missing | Pages linked |
|---|---|---|---|---|
| Best brunch café in [town] | ||||
| Cost of [job] in [town] |
Then sort the linked pages by kind. A "best of" list, a directory or a review site is somewhere you may be able to get listed. A competitor's own page, written for the question you're being asked, is a page you probably need your own version of. A forum thread is something to read and learn from. Never post in one pretending to be a customer.
Step 2: let PerplexityBot read your site
Perplexity runs two bots, and Perplexity's own documentation describes them differently:
| Bot | What Perplexity says it does | What blocking it means |
|---|---|---|
| PerplexityBot | Surfaces websites in Perplexity search results. Perplexity says it is not used to crawl content for AI foundation models. It respects robots.txt. | Your pages are less likely to be surfaced in Perplexity search results. |
| Perplexity-User | Visits a page when a person's question needs it. Perplexity says it generally ignores robots.txt, because a user triggered the visit. | A robots.txt rule is not a reliable way to stop it. Use your host's bot settings if you really need to. |
Check two places. First, open yourwebsite.com/robots.txt and look for User-agent: PerplexityBot followed by Disallow: /. If there's no group for it, it follows User-agent: *. A file that welcomes everyone is:
User-agent: * Allow: / Sitemap: https://www.example.com/sitemap.xml
Second, check your host or security service. A robots.txt file only tells bots what you'd like. Firewalls, Cloudflare's bot settings, Wordfence, Sucuri and "bot protection" on some hosts can block a crawler no matter what robots.txt says. Perplexity's documentation says both bots publish their IP addresses (PerplexityBot, Perplexity-User) and recommends using both user-agent matching and IP checks in a firewall rule, so a rule that allows the real bots does not also let a faker through. Your web developer can do this in a few minutes.
Whether to allow training bots from other companies is your call and a separate one. Blocking a training bot doesn't by itself remove you from an answer engine, which is covered in the ChatGPT guide's crawler step. Then view your page with "view source": if your prices and hours aren't in the HTML, a crawler that doesn't run JavaScript won't see them.
Step 3: put the answer in the first lines of the page
Perplexity's answers are short, and each claim is tied to a link. A page is easy to cite when the sentence that answers the question is right there, with the numbers. For each thing you want to be named for, have one page whose first paragraph says what it is, who it's for, where, and what it costs or how it works.
| Hard to cite | Easy to cite |
|---|---|
| Welcome to our family-run plumbing business! We pride ourselves on service and care about every customer. | Example Plumbing fixes blocked drains in Penrith and the Blue Mountains. A standard blocked drain is $180 to $350 depending on access, and we can usually attend the same day. |
Example only: the business and the prices are invented for this guide.
- Use the questions from Step 1 as headings, and answer each in the sentence below.
- Show a price or range, a service area and what's included. Cost questions are common, and a page with no number gives Perplexity nothing to quote.
- Keep it as text. An answer inside an image, a slider or a PDF is easy to miss.
- Date the facts you checked. "Prices checked 14 October 2026" helps people and machines. Only write it if you've checked.
- Add a short FAQ with the questions customers ask you on the phone.
The full version of this step, with schema and a fact sheet, is in the ChatGPT guide, and what AEO is explains why answer-first pages get picked.
Step 4: make your facts match everywhere
If your name, address, phone, hours and prices differ between your site, Google Business Profile, Bing Places, Facebook and directories, an answer can quote whichever one it happened to read. Write one fact sheet and make every place match it. Fix old listings too: a directory page with a dead phone number is just the sort of page that gets cited. Step 2 of the ChatGPT guide has the checklist.
Step 5: get onto the pages Perplexity already links to
Go back to the pages you wrote down in Step 1. Perplexity shows its sources, so you can check whether the cited lists and directories include your business, and seek inclusion where relevant.
- Count which pages are linked most across your questions, and sort them: directories you can claim, lists written by someone you can contact, review sites, forums.
- Claim and complete every free listing on that list, with the same facts as your fact sheet.
- Contact the writers of the lists you belong on. Be short and honest, and bring something worth including: a new menu, a real story, an offer. Don't pay for a link that passes ranking credit, or for a sponsored spot dressed up as independent.
- Earn reviews honestly. Invite every customer, leave the wording to them, reply to every review, and never buy, write or reward them.
- Read forum threads that come up, to learn what people ask and how they talk. If you take part, say who you are.
Step 6: ask again and compare
Repeat the Step 1 questions each month, the same ones, a few times each, and count. "Named in 3 of 12 questions that don't mention us" is a result; "it mentioned us once" isn't. A change of one or two either way is noise. Also check your analytics for visits from perplexity.ai. Many AI-sent visits arrive unlabelled, so you'll undercount, but a rise is a good sign.
There is no established timeline for earning citations. Perplexity hasn't said how fast it re-reads a site, so treat any promised timeline with suspicion and repeat the same questions monthly.
What doesn't work
- Paying for a guaranteed mention. No one controls Perplexity's answers.
- Hidden text aimed at AI, or serving crawlers a different page than people see.
- Hundreds of near-identical suburb pages. They give an answer nothing new to cite.
- Fake reviews and fake forum posts.
- Judging by one answer. Answers differ between asks.
The one-page checklist
Tick them off as you go. The print button prints this list on its own.
- Ask the five questions from Step 1 on Perplexity, a few times each, and save the linked pages.
- Sort the linked pages into directories, lists, reviews, competitors and forums.
- Check robots.txt for PerplexityBot, and your host's bot settings.
- Check that prices and hours are in the page HTML, not only in images or scripts.
- Rewrite your main pages so the first paragraph says what, who, where and the price.
- Write one fact sheet and make your site, Google Business Profile, Bing Places and directories match it.
- Claim every free listing on the pages Perplexity linked.
- Invite honest reviews and reply to every one.
- Repeat the questions monthly and count how often you're named.
Questions
How do I rank on Perplexity?
To become one of the pages Perplexity cites, let PerplexityBot read your site, answer the buyer's question clearly, keep your facts consistent and seek inclusion on relevant cited pages.
Can I pay to be recommended by Perplexity?
Not for the answer itself. Be wary of anyone selling a guaranteed spot in what it writes.
What is PerplexityBot?
Perplexity's crawler for surfacing websites in its search results. Perplexity's documentation says it is not used to crawl content for AI foundation models, and that it respects robots.txt.
Does Perplexity-User follow robots.txt?
Perplexity says it generally ignores robots.txt rules, because a person's question triggered the visit. Use your host's bot settings if you need to restrict it.
Is ranking on Perplexity the same as ranking in ChatGPT?
The method is mostly the same: readable pages, matching facts, answer-first writing and mentions on trusted pages. The crawler names differ, and Perplexity shows its sources more openly, which makes Step 1 more useful. See the ChatGPT guide.
How long does it take?
There is no established timeline for earning citations. Perplexity hasn't published how often it re-reads sites, so repeat the same questions monthly and compare.
Will this work for my kind of business?
Yes. The steps are the same for a café, a plumber, a dentist or a shop. What changes is the questions your customers ask, so spend your time on Step 1 for your own trade.
Do it yourself, or have Dinkum do it every weekday morning
Everything here you can do yourself, and we'd rather you did than did nothing. The hard part is keeping it going: asking the questions every week, noticing which answer got something wrong, rewriting the page that should have won.
That's what Dinkum is. It asks a fixed set of buyer questions each week across ChatGPT, Perplexity and Google's AI answers, finds the page or fact that needs fixing, and ships one change to your site each weekday as a pull request you approve, then judges each change at 14 and 28 days against the rest of your site. It needs access to your website's code, because that's how it makes the changes, and it costs $2,000 a month for one site, month to month. There are no published client results yet, so we won't quote any. See how Dinkum works, and what an AEO agency should do for what to expect from anyone you hire.