Assistant · Perplexity AI

Perplexity SEO for SaaS

Perplexity SEO starts here. Small by usage and disproportionately useful, because it is the one assistant that shows its working: sources are listed on screen, so citation share is directly observable rather than inferred.

2%of US assistant usage

Source: First Page Sage, Top Generative AI Chatbots by Market Share. July 2026, US monthly-active-user estimates.

A marketing lead reviewing results at a standing deskAssistant · Perplexity AI
Retrieval mechanism

Search-first, with the sources on screen

Perplexity searches for essentially every query rather than answering from training memory. There is no meaningful training-mode failure to diagnose here: if you are not in the answer, you were not retrieved, and that is a much more tractable problem.

It also displays the sources it used, numbered, next to the answer.
Every other assistant makes you infer citation behaviour from sampling. Perplexity shows it, which makes it the best available instrument for understanding what the whole category rewards, regardless of its own market share.
For that reason we treat Perplexity as a diagnostic surface first and an audience second.
Two percent of assistant usage does not justify a large standalone programme. Being able to watch which sources get cited for your category, in the open, usually does.
Two colleagues reviewing work on a laptop
Crawler policy

Which crawler does what, and what blocking it costs you

  • Training crawlers and retrieval crawlers are different things
  • Blocking a retrieval crawler removes you from the answers
  • Blocking a training crawler does not
RetrievalPerplexityBot
Builds the index Perplexity searches. Blocking it removes you from the source list entirely, which on this platform is the whole of your Perplexity visibility.
User-triggered fetchPerplexity-User
Fetches a page when a user follows a citation.

Our own policy is published at surge45.com/robots.txt, with a line per agent and the reasoning in the file.

A growth lead planning the next quarter
Citation behaviour

How the source list is built

Perplexity retrieves a set of pages, writes an answer from them, and lists them numbered beside the text. Position in the list is visible and clickable, which makes it the only assistant where referral traffic from being cited is directly measurable in your own analytics.

The retrieved set favours pages that answer the specific question asked.
Broad category pages perform poorly against narrow pages that address the exact query, which is close to the opposite of how a conventional search authority calculation behaves.
Follow-up questions re-retrieve.
A source cited on the first answer frequently disappears on the second, so single-prompt measurement systematically overstates how well you are doing.
Two colleagues talking at a standing desk
What moves it

Observed to change Perplexity visibility

01Narrow pages that answer one question completely
Specificity beats authority here more clearly than on any other platform. A focused page on a single buyer question outperforms a comprehensive hub.
02Being present on the domains it retrieves
Review platforms and community threads appear constantly in Perplexity source lists for software queries. The visible list makes it easy to see exactly which ones, for your category.
03Clean, crawlable, fast pages
Retrieval happens at query time. A page that is slow, JavaScript-dependent or blocked simply does not make the list.
04Using it as the measurement instrument
Run your prompt set here first. The visible source list tells you which domains own your category's citations, which then directs the work on every other platform.
Two colleagues working through a plan on a laptop
What does not

Sold for Perplexity, and not worth buying

✕ Domain authority plays
Retrieval selects for question-answer fit, not accumulated authority. Large sites lose to specific pages here routinely.
✕ Treating it as a volume channel
Two percent of assistant usage. Its value is as an instrument and as a high-intent, low-volume referral source, and we would rather say that than sell it as a growth channel.
A team looking at results on a monitor together
Working together

What we do on Perplexity

  1. /01
    Run the full prompt set and record the source list for every answer, including follow-ups.
  2. /02
    Build the map of which domains own citations in your category, from observed data rather than assumption.
  3. /03
    Use that map to direct the citation-sources programme across every platform.
  4. /04
    Fix crawlability and speed on the pages that should be retrieved and are not.
  5. /05
    Track referral traffic from Perplexity citations as a real, attributable number.

Perplexity questions we get asked

Why work on a platform with two percent share?

Mostly because it shows its sources. It is the only place you can watch, rather than infer, which domains get cited for your category, and that observation directs the work everywhere else. As an audience it is small and high-intent. We would not sell it as a volume channel.

Can we see traffic from Perplexity citations?

Yes, and it is one of the few AI surfaces where you can. Citations are clickable links, so referrals appear in your analytics as an attributable source rather than as an inference.

Who helps SaaS companies optimise for Perplexity?

Usually a GEO agency rather than a dedicated Perplexity SEO agency, because Perplexity's main value is as an instrument for every platform. Surge45 runs Perplexity AI visibility optimisation as part of GEO: we run your prompt set, record the source list for every answer and use it to direct the work everywhere else, alongside your team.

Ready when you are

Ready to build your GTM engine?

Book a discovery call with a GTM engineer. We'll identify which tier fits your stage and which part of the engine to build first.

A smiling strategist standing in front of a whiteboard of notes