Perplexity visibility and citation tracker

Perplexity is built around direct answers with linked sources. That makes it a useful surface for understanding both sides of brand visibility: which companies are named in the response and which domains supply the evidence.

refd keeps those signals separate. A source can influence an answer without the source owner’s brand being named, and a brand can be discussed without its own domain being cited.

What Perplexity visibility means

Perplexity’s current product documentation describes the service as an AI-powered search engine that searches the web, produces conversational answers, and links to original sources. Perplexity also offers several search modes and, for some plans, a choice of underlying model.

refd tracks the configured buyer questions in Perplexity and records the visible answer text and source URLs returned with each answer. The tracked brand receives citation credit only when a normalized URL matches one of its confirmed domains.

refd does not infer ownership from a page title, guess a brand from an opaque redirect, or count a source-card title as a visible mention.

What refd measures on Perplexity

Signal	What it tells you
Mention rate	How often Perplexity names the brand in answered prompt and sample cells
Citation rate	How often a returned source URL belongs to the brand
Average position	Where the brand is first named relative to tracked competitors
Share of voice	The brand’s share of tracked-entity mentions or citations
Source domains	Which owned and third-party domains support relevant answers
Prominence	Whether the strongest mention is in the lead, body, or a list
Sentiment	How the answer describes the brand when it is mentioned

Mention rate and citation rate use the same eligible answer set and equal cell weighting. An answered result with no brand citation remains a zero for citation visibility. Review the methodology for the exact denominator and URL normalization rules.

Read the illustrative result

The fabricated sample asks which wearable is best for metabolic health insights. Perplexity names Ultrahuman first, connects the product to metabolic context, and cites ultrahuman.com.

The answer is valuable for more than a positive tile:

• Position 1 shows that Ultrahuman led the tracked competitive set.
• The brand-domain citation shows direct source visibility.
• Other cited domains show which adjacent products or publishers shaped the response.
• The raw prose explains which attributes created the recommendation.

If the citation disappeared in a later run while the mention remained, that would be a source change rather than a loss of all visibility. refd preserves both signals so the difference remains visible.

Use source gaps as research input

Perplexity source gaps are domains cited on relevant answers where your brand was neither mentioned nor cited. They can reveal:

• Comparison pages that define the category.
• Publisher reviews that support competitor recommendations.
• Documentation pages used to explain technical criteria.
• Community discussions that supply first-hand experience.
• Competitor pages that own a specific attribute or use case.

A source gap is not an instruction to copy or manipulate a page. Read the source, understand why it was relevant, and decide whether your own product, documentation, evidence, or positioning leaves a genuine information gap.

Build a Perplexity prompt set

Include questions from several buyer stages:

1. Discovery questions that describe the problem without naming a vendor.
2. Category comparisons that ask for a shortlist.
3. Alternative questions around a known competitor.
4. Evaluation questions about requirements, integrations, cost model, or limitations.
5. Decision questions for a specific company type or use case.

Avoid filling the set with slight rewrites of one keyword. A broad but stable question set produces a more useful picture of what Perplexity says across the buyer journey.

Limitations

Perplexity answers can change between identical prompts. Search mode, selected model, available sources, location, session context, and product updates may affect both prose and citations. The hosted tracker records a defined observation, not every possible Perplexity experience.

Although Perplexity describes its answers as cited, refd reports only the source URLs actually returned with the answer. A missing URL remains missing evidence. refd never manufactures a citation to make the result look complete.

Frequently asked questions

Does refd count every Perplexity source as a brand citation?

No. A source is attributed to a brand only when its normalized hostname matches a confirmed domain for that tracked entity. Other URLs remain third-party sources.

Why can citation rate differ from mention rate?

The answer might name a brand but cite a publisher, or it might cite a brand page without naming that brand in the visible prose. These are different signals and should not be collapsed.

Does refd control which Perplexity model produces the answer?

No. Model selection is not a workspace control. refd records the tracked result and does not claim that one observation represents every Perplexity mode or plan.

Sources reviewed

• What is Perplexity?, Perplexity Help Center
• How refd measures AI search visibility

External product behavior was reviewed on 30 July 2026. The refd collection and metric descriptions match the open-source implementation on that date.