I'll tell you the pattern we kept seeing before we understood it. A client would be crushing it on Google — top three for half their keywords, solid domain authority, hundreds of citations. And Perplexity would cite someone we'd never heard of. A smaller competitor with a cleaner story.
At first we blamed the algorithms. Then we looked closer at what the cited sites were doing, and the reasons turned out to be depressingly concrete.
Understanding Perplexity's source selection
Perplexity is fundamentally different from Google in how it selects sources. Google ranks pages based on hundreds of signals and presents a list for users to evaluate. Perplexity constructs an answer from the sources it trusts most, then cites those sources directly. The selection process prioritizes clarity, consistency, and verifiability over traditional ranking signals.
This means a page that ranks #1 on Google might not be Perplexity's top choice if a competitor has clearer, more extractable content. Perplexity isn't looking for the "best" page — it's looking for the most defensible source to quote.
For a broader look at how different AI engines select sources, see our guide on what AI visibility is and how to measure it.
Reason one: Your facts aren't quotable
It's the top reason pages fail in our AI Visibility Alerts scans, and it's fixable in an afternoon:
Perplexity builds an answer and then hunts for a source it can point to. It wants a sentence that says the thing. If your page answers the question by implication, by vague positioning, or across six scattered paragraphs, there's nothing clean to cite.
We've had pages where the exact answer to the buyer's question existed — but buried in a two-paragraph run-on at the bottom of an otherwise generic page. Perplexity skipped it because it couldn't extract it quickly.
Fix: put the direct answer in the first two sentences. If the query is "how much does X cost," your first line should literally state a number, not promise to reveal one later. Write answers the way you'd want an AI to quote them: clear, specific, and extractable.
Reason two: Your identity isn't machine-readable
This is the one that surprises business owners most. Perplexity reads your HTML like a researcher skimming a messy case file. If you don't have proper schema, it can't cleanly attribute the statement to your business as an entity.
We added Organization, Service, and FAQPage structured data across several sites and watched citation counts rise even though the visible text hadn't changed. The AI could finally tell who you were and what you claimed.
Our guide on structured data for AI visibility covers the specific schema types that matter most for Perplexity citations.
Reason three: Nobody corroborates you
Perplexity doesn't trust a single page. It tends to prefer facts that appear consistently across multiple references — your own pages, third-party mentions, reviews, case studies. If your key claim exists only in one place, it reads as unsupported.
Fix: this is the brand-consistency work I keep coming back to. Make your number, your differentiator, and your process repeat identically everywhere your brand appears. Our guide on AI trust signals explains the specific signals that build corroboration.
Reason four: You have nothing original to cite
Here's the uncomfortable one. Perplexity has access to a million roundups that all say the same thing. If your page is another generic listicle, why would it pick you? It picks the source with an actual point of view, original data, or a specific experience.
The sites that win citations are the ones that produce something only they could have produced — the survey you ran, the method you named, the example you lived.
Reason five: Your content isn't structured for extraction
Perplexity favors content that's easy to extract and quote. This means:
- Clear headings that signal what each section covers
- Direct answers in opening paragraphs
- Specific numbers and facts rather than vague claims
- Concise paragraphs that can be quoted without losing context
- FAQ sections that directly answer common questions
Our guide on content that ranks in AI search covers content structure tactics for AI engines.
Reason six: Your cross-web presence is weak
Perplexity evaluates your brand across the entire web, not just your website. If your Google Business Profile is incomplete, your directory listings are inconsistent, and your social media presence is sparse, Perplexity has less evidence to trust you as an entity.
Building a strong cross-web presence means:
- Complete and accurate Google Business Profile
- Consistent business information across all directories
- Active social media profiles that reinforce your entity
- Reviews and testimonials from real clients
- Mentions in industry publications or roundups
Our guide on entity SEO for AI search covers building comprehensive entity signals.
The baseline: counting your citations before you change anything
You can't tell whether your fixes are working unless you establish a baseline first. Here's the method we use with every client:
- Pick ten buyer-intent queries — the questions your customers actually ask. Not brand keywords; plain-language questions.
- Ask each one in Perplexity. For every answer, record whether your site or brand is mentioned, how it's described, and whether any source link points to your site.
- Repeat with your top three competitors so you have a comparison, not just an absolute number.
- Record the date and the answers. Screenshots or a simple spreadsheet both work.
- Re-run the same ten queries monthly and count. This is the only honest way to measure progress.
The point isn't to get a perfect score on one day. It's to catch the movement when your fixes land. The pitfall most people hit is skipping the baseline and then guessing whether anything changed.
What a winning page looks like: before and after
Here's a concrete example of the shape most cited pages take. You don't need the full rewrite to see the difference.
Before (ignored): "Acme Landscaping is a full-service landscaping company serving the greater area. We believe quality matters and we take pride in our work. Contact us today for a free quote!" — no facts, no numbers, no reason to cite.
After (cited): "Acme Landscaping installs interlocking patios in Ottawa and averages 3-4 weeks from design to completion. A typical 300 sq ft patio runs $8,000-$12,000 including materials and removal. We offer a two-year workmanship warranty." — facts, prices, timelines, and a location. An engine can quote every sentence.
Notice the second version never got worse for humans. It's the same page with the preamble stripped out and the real information moved to the top.
The reason-and-fix map
Here's the whole diagnosis on one page, in the order you should work through it:
| Reason | The fix | When you see results |
|---|---|---|
| Facts aren't quotable | Answer directly in the first two sentences | Fast — weeks |
| Identity isn't machine-readable | Add Organization, Service, FAQPage schema | Fast — weeks |
| Nobody corroborates you | Standardize facts across every source | Medium — a month or two |
| Nothing original to cite | Publish data, opinions, or named methods | Slow but compounding |
| Poor structure | Headings, concise paragraphs, FAQ blocks | Fast — weeks |
| Weak cross-web presence | Complete GBP, consistent listings, reviews | Medium to slow |
That ordering is deliberate: the fixes at the top are cheap and fast, which makes them the most likely to get done. The slow ones are the ones that build the entity, and they compound over time.
What not to waste effort on
A few things come up repeatedly that won't move Perplexity citations, no matter how well you execute them:
- Aggressive link building. Directory spams and paid links don't add authority for an answer engine. Kill two birds: invest in facts and corroboration instead.
- Publishing volume over originality. Twenty more pages of the same advice just give the engine twenty more reasons to prefer a single strong source.
- Obsessing over Google position. A #1 Google ranking is useful, but it measures a different game. Perplexity is checking whether your page is the most defensible quote for the answer, not the best-ranked result for the keyword.
- Chasing every engine at once. Optimize for the two engines that actually send you traffic. The rest will largely follow from the same entity work.
The practical fix list
- Answer the question in the first two sentences, plainly.
- Ship the schema that names you as an entity.
- Sweep the site for contradictions and make them agree.
- Publish at least one genuinely original, opinionated piece per month that states your position out loud.
- Re-run your baseline queries in Perplexity and count enumerable citations, not vibes.
The good news is Perplexity recrawls frequently, so well-aimed fixes show up faster than you'd think. The bad news is the generic roundup approach will never win, because the AI doesn't need another generic roundup.
For a deeper dive, see our guide on running an AI visibility audit.
For a deeper dive, see our guide on understanding your AI visibility score.
We help brands turn from "mentioned" into "cited." If you want to see your own citation baseline, book a free website teardown and we'll run the numbers.