Blog

How to Improve Brand Visibility in Perplexity

Robert VîjaRobert Vîja
November 16, 20268 min read
How to Improve Brand Visibility in Perplexity

Perplexity builds most of its answers from concrete pages on company sites and from press coverage it already trusts. It doesn’t hand out a ranked list of links. Brand visibility in Perplexity comes down to owning a page for every buyer question, and showing up in the comparison and news pages Perplexity already cites.

Marketers keep asking me the same question: why does my brand disappear from Perplexity when it ranks well on Google? This article covers how Perplexity picks sources, what makes a page worth citing, and how to check your progress.

How to improve brand visibility in Perplexity

Start with your own pages. Perplexity mostly cites concrete pages on company sites, plus press coverage.

At GEOflux we run a public State of AI study. From 18 October to 16 November 2026, it logged about 42,000 Perplexity answers and 519,000 citations. The prompts were commercial questions in banking, insurance, fashion, e-commerce and travel, across Romania, the UK, Spain and Germany.

The pattern is clear. 56% of citations go to company sites. 28% go to the tracked brands’ own sites. So more than half of Perplexity’s sources are company websites, and over a quarter of the time it’s the exact brand being asked about.

Bar chart: share of citations, Perplexity vs ChatGPT. Company sites 55.6% vs 63.7%, media and blogs 14.9% vs 5.8%, marketplaces and retailers 14.8% vs 15.8%, review and comparison sites 8.6% vs 7.1%, Reddit 0.7% vs 0%.

That changes the playbook. You don’t win Perplexity by chasing rankings. You win it by owning the page that answers a buyer’s exact question, and by appearing in the coverage Perplexity already relies on.

How does Perplexity AI choose sources to cite

Perplexity pulls in a lot of pages per answer. And it prefers pages that answer the question directly over general brand pages.

On the prompts in our State of AI study, Perplexity cited more sources per answer than ChatGPT or Gemini, and close to what Google AI Mode pulls in.

AI engine Average sources cited per answer
Google AI Mode 12.5
Perplexity 12.3
ChatGPT 8.5
Gemini 3.7

The type of page matters more than the count. We classified Perplexity’s citations by URL and title, so treat this split as indicative:

  • Concrete commercial pages (product, category, service): 54%
  • Comparison, top-list and review pages: 31%
  • Articles and guides: 7%
  • Homepages: 4%
  • Forums and video: under 1%

The homepage number surprises people. Perplexity almost never cites a homepage. It wants the page with the answer: a loan rate, a product spec, a service area.

ChatGPT works differently. About half of its links go to a brand’s homepage. If you built your pages around ChatGPT’s habits, Perplexity needs a different, page-level approach. I cover ChatGPT’s side in how to rank in ChatGPT.

What factors influence Perplexity AI answers

Press coverage and fresh pages matter more than raw domain size.

Media and blogs are 15% of Perplexity’s citations, against 6% in ChatGPT. Perplexity leans on independent reporting much more.

Review and comparison sites are 9% of citations overall. But that average hides a big gap by country: 4% in Romania, 19% in the UK.

Here’s the surprise: Reddit barely shows up. It’s 0.7% of Perplexity citations on our commercial prompts. YouTube is practically zero, with 9 citations out of 519,000. That contradicts the common advice that Perplexity leans on Reddit. How-to or technical questions may look different. For banking, insurance, fashion, e-commerce and travel, Reddit isn’t the lever people think it is.

Freshness matters too. We don’t store update dates, so we use a proxy: cited pages with a year in the title or URL, which are 18% of citations. Of those, 93% carry the current or next year. Only 3% carry last year. It’s the same on all four engines we checked.

In practice, a page still titled “2025” is almost never cited in November 2026. Keep dated pages current, or take the year out of the title.

How to rank on Perplexity

There’s no rank on Perplexity. What you’re after is presence across repeated answers.

Answers move every day. Same prompt, same persona, consecutive days: Perplexity keeps 41% of the brands it named the day before. ChatGPT keeps 49%. The first brand named stays the same in 47% of cases on Perplexity, and 57% on ChatGPT.

Sources behave the opposite way. 36% of the URLs Perplexity cites repeat the next day, against 21% in ChatGPT. Perplexity keeps going back to the same pages, even when the brand list shuffles.

So one answer tells you nothing. A brand can vanish today and come back tomorrow on the same prompt. Judge visibility on averages over several days.

Bar chart: same prompt on consecutive days, Perplexity vs ChatGPT. Brands kept the next day 41% vs 49%, same first-named brand 47% vs 57%, cited URLs repeated 36% vs 21%.

What does “ranking” mean then? Build one page per buyer question, readable by PerplexityBot. Then get your brand into the press and comparison pages Perplexity cites for the same question. Own both the source and the mentions, and you keep showing up while individual answers shuffle.

Improving website authority for Perplexity AI

In Perplexity, authority shows up as your product pages getting cited more often. It isn’t a score you can look up.

Here’s an example from our public State of AI study. It’s not a client, so I won’t name it. A Romanian bank went from 42% to 53% of Perplexity answers, comparing one 30-day window with the one before. Week by week the jump was bigger: from 43% to 69%.

Over the same period, the bank gained only 5 points in ChatGPT and 1 point in Gemini. Whatever happened was specific to Perplexity.

The data shows what changed. Perplexity started citing the bank’s product pages (loans, deposit rates) more than twice as often. Total citations in the study didn’t grow. Perplexity just moved its attention to this bank’s answer pages.

Honest answer: we don’t know what the bank changed on its site. We only see the before and after in Perplexity. What we know is that the shift was real, specific to Perplexity, and tied to product-page citations.

Optimizing content structure for Perplexity citations

Structure decides whether Perplexity can lift your content into an answer.

My rule: one page answers one buyer question, completely, in the first lines. If someone asks what a loan costs, give the rate, the term and the conditions, near the top.

The page also needs to read without JavaScript. Don’t assume a crawler runs your scripts. If your rates or specs only appear after a script loads, they may never be seen. I’d take a plain page that loads as HTML over a beautiful one a bot can’t read.

In my view, schema helps too. It delivers the same facts in a form that’s easy to parse, much like rich snippets did for classic search. It doesn’t replace clear writing. It backs it up. I go deeper into this in our piece on structured data for AI search.

Put together: one question, real figures, schema, readable without JavaScript. That’s the page Perplexity is built to cite.

How to get cited by Perplexity AI

Find the third-party pages Perplexity already trusts for your buyers’ questions. Then work to get your brand into those pages.

Press and blogs are 15% of Perplexity’s citations, almost three times ChatGPT’s share. Perplexity leans on independent coverage.

The step is simple to describe. For each question your buyers ask, run it in Perplexity. Note which press articles and comparison pages it cites. Those are the pages Perplexity trusts for that question.

Then it’s relationship work. Pitch those publications. Get into their comparison pages. Make sure your product and rate data is accurate enough that a journalist wants to quote it.

There’s no shortcut I can promise. What I can tell you is where to look: the pages already cited for your buyers’ questions are your target list. A random list of big domains is not.

Perplexity AI brand mentions optimization

Mentions and citations aren’t the same thing. The gap between them is where the work is.

When Perplexity names a brand we track, it also cites a page of that brand in the same answer 81.5% of the time. Only 18.5% are mentions without a link.

AI engine Brand mentions that come with a citation of the brand’s site
Perplexity 81.5%
ChatGPT 76%
Google AI Mode 59%
Gemini 17%
Bar chart: brand mentions that come with a citation of the brand's own site. Perplexity 81.5%, ChatGPT 76%, Google AI Mode 59%, Gemini 17%.

Gemini is the outlier. Most of the time it names brands without citing them.

The reverse happens too. In about one fifth of the cases where Perplexity cites a brand’s site, it doesn’t name the brand in the text.

My view: the mention is what the reader sees. The citation is the lever you can actually work on. You can’t make Perplexity say your name. You can build the page that earns the citation, and the mention often follows.

How to track brand mentions in Perplexity

Run the same prompts on a schedule and watch how answers and citations move.

That’s what our Perplexity visibility tracking does. GEOflux runs your prompts daily, per persona, and shows visibility, mentions, share of voice and sentiment, down to the exact URLs Perplexity cites for each prompt.

You can put any domain in a watchlist, including competitors or the press sites above, with daily, weekly or monthly email reports.

And can Perplexity reach your pages at all? Our AI Readiness check looks at whether PerplexityBot and Perplexity-User can access the site (robots.txt, CDN, firewall) and what a page shows without JavaScript.

Alerts aren’t Perplexity-only. They measure your brand across all the engines you track, so you see Perplexity next to ChatGPT, Gemini and the rest.

How to increase brand citations in AI search results

Once Perplexity reads your own pages regularly, citations compound.

Blocking Perplexity’s crawler doesn’t make you disappear, but it costs you control. Of 116 large brand sites we checked, 4 block PerplexityBot. A large retailer that blocks it is still named in 25% to 60% of Perplexity’s e-commerce answers. Perplexity just describes it from other people’s pages.

What separates brands is how many of their own pages Perplexity reads. In the same study, a retailer with more than 3,000 citations to its own pages appears in 90% of answers. One with under 300 appears in 22%.

Do this week: take the 10 questions your buyers ask most. Run each one in Perplexity. Write down the cited URLs. Then check two things: does your site have a page that answers each question exactly, and can PerplexityBot read it?

Share
Robert Vîja

Robert Vîja, Co-founder & CPO @geoflux.ai

Runs product at GEOflux - what gets tracked, how citations are measured, what ships next. Ten years in SEO, on projects above 1M organic visits a month.

Ready to see your AI visibility?

Create a free account and discover how AI perceives your brand in ChatGPT, Gemini and Google AI Mode. No card, no time limit.

Start for free