What Perplexity Reads When It Recommends a Business

When someone asks Perplexity "best accountants in Hamilton" or "top AEO audit tools," it does not visit a roster of business websites and evaluate them. It finds pages that have already done the comparison work -- category roundups, directory listings, tool review sites -- and synthesizes its recommendation from those.

That's the mechanism. And it's different from what most businesses are optimizing for.

What our source analysis found

In our May 2026 investigation of platform citation behaviors (session 16, 2026-05-13), Scout systematically analyzed every source URL Perplexity cited in a live Sourcepull audit -- 102 URLs total across category, brand, and service queries. This is the first time we mapped Perplexity's actual source documents empirically, rather than inferring from scores.

The category query results were the sharpest departure from conventional wisdom.

**Reddit: completely absent.** Zero Reddit URLs across all 102 captured sources. In prior years, Reddit dominated Perplexity's category recommendations -- it was the community discussion layer Perplexity relied on for "what do real users think?" answers. After Reddit sued Perplexity for unauthorized scraping in October 2025, Perplexity's Reddit citation share dropped approximately 86%. As of May 2026, it has not meaningfully recovered. For the category we analyzed (AEO audit tools), Reddit is not a Perplexity source at all.

**A tiny exact-match domain appeared as often as a well-funded competitor.** A site called aeoaudittool.com -- a small independent tool directory with no apparent press coverage and no significant domain authority -- appeared in 3 out of 4 "best AEO audit tools" queries. That is identical frequency to tryprofound.com, a well-funded enterprise competitor with a full content program. On traditional domain authority logic, there is no reason these two domains should appear at the same rate.

**The G2 category page appeared in 3 of 4 queries.** Not individual product listing pages within G2, but the category-level page that aggregates all tools in the AEO category. Perplexity is reading the comparison layer, not your specific listing.

Why an unknown site outperforms established competitors

Our May 14 investigation into aeoaudittool.com found the explanation: the domain name is a keyword-exact match for the query concept.

Perplexity's live retrieval layer favors pages where the URL and page title directly signal category relevance. A domain called `aeoaudittool.com` covering AEO audit tools is, from Perplexity's perspective, almost unambiguous about what it covers. It faces minimal competition for that category signal regardless of how many backlinks it has or how long it has existed.

From our investigation (2026-05-14): "A page titled 'best AEO audit tools' on a $9/year domain can outperform an established media site for the 'best AEO audit tool' query if the domain and title are direct keyword matches."

This is not a quirk of the AEO category. The mechanism generalizes. A local site called `bestplumbersburlington.ca` -- even with 15 posts and no inbound links -- is a Perplexity citation candidate for Burlington plumber category queries because the domain signals categorical intent. A large agency site with a single buried service page competes poorly against it for that specific query type.

The implication: for Perplexity category visibility, the question to ask is not "how do I build more domain authority?" It is "which pages specifically about my category is Perplexity already reading? Am I mentioned in them?"

The G2 category page problem

The finding about G2 category pages is worth unpacking separately, because it changes what "being on G2" actually means for Perplexity.

Most businesses with a G2 presence focus on reviews and their product listing page. That is where social proof accumulates, where conversion happens for human visitors. But our source analysis shows Perplexity is reading `g2.com/categories/[your-category]` -- the aggregated category page -- not individual product pages.

The category page is what gets cited when someone asks Perplexity "what are the best [category] tools?" Getting listed in the correct G2 category is the step that matters for Perplexity. Being on the platform at all is not sufficient if you are miscategorized or absent from the category page Perplexity is reading.

This pattern almost certainly applies to Yelp, Capterra, Houzz, HomeAdvisor, and other category-aggregating directories. The category page -- "best dentists in Hamilton," "top rated contractors in Calgary" -- is likely what Perplexity cites, not the individual business listing within it. Being listed on the directory gets you onto that page. Being absent from the directory keeps you off it.

How this compares to ChatGPT and Gemini

This source architecture is Perplexity-specific. Our platform citation behaviors investigation (session 15, 2026-05-12) documented the contrast using Yext's 2026 AI Visibility study -- 17.2M citations across ChatGPT, Perplexity, and Gemini.

ChatGPT pulls 49% of its citations from third-party directories. The individual listing -- your Yelp page, your BBB entry, your G2 product listing -- is what feeds ChatGPT. It is reading the business's specific entry, not the category page containing all businesses.

Gemini sources 52% of citations from brand-owned websites. Gemini is reading your site more directly than the other platforms. Schema markup, E-E-A-T signals, and Google Business Profile carry higher relative weight on Gemini than on Perplexity.

Perplexity specifically trusts niche expert directories and category-intent pages -- the comparison layer that sits above individual business listings. The Yext data documents this directional difference; our source_url analysis gives it empirical backing from actual audit data.

Same category query, three different source architectures. A fix that moves ChatGPT citations often has limited effect on Perplexity, and vice versa.

What to actually do about Perplexity category visibility

The Perplexity category fix is not more schema markup or more blog posts. The investigation starts with: which comparison articles, roundup pages, and directory category pages exist for your business type and geography? Which of those have keyword-matching URLs and titles? Those are the pages Perplexity is reading. If your business is not named in them, the fix is getting into those pages -- not improving your own site's authority.

For most service businesses, this means: verify you are listed in every major directory that operates category pages for your niche (Yelp, HomeAdvisor, Houzz, Avvo, Healthgrades -- whichever apply), and ensure your business name appears correctly on the category aggregation pages, not just in your individual listing. For SaaS or tool businesses, it means getting into category listings on G2, Capterra, and any niche comparison directories that Perplexity is actually reading for your category.

The good news: reaching these pages is often more accessible than building traditional domain authority. Perplexity's source selection creates openings for newer and smaller businesses that do not exist in traditional search rankings.

A full Sourcepull audit captures the actual source URLs Perplexity cites per query type -- including which category pages it's currently reading when evaluating your business. That's the fastest way to know which specific comparison articles and directories to target. If you want to start with a quick look at your current Perplexity score alongside ChatGPT, Gemini, and Claude, the Signal Check at sourcepull.ca takes about 60 seconds.