Skip to content
FRFree GEO audit

Observatory · Engine

Which sources does Perplexity cite on buying questions?

Perplexity cites 9.7 sources per answer on buying questions, the highest of the six engines in the Observatory, and 33.4% of those sources are rankings and comparison pages. Data: 106,045 answers collected via Profound between July 12 and October 11, 2026.

Get my free GEO auditBook a 30-min call

Updated

92%

answers citing sources

9.7

sources per answer

33%

rankings among sources

64%

first-cited brand changes week to week

Types of sources cited : Perplexity
Other sites60%Rankings and comparisons33%Reviews and directories5%

Data : Profound, 106,045 answers analysed, July 12, 2026 to October 11, 2026.

Share of rankings among cited sources, by sector (Perplexity)
Marketing agencies57%Finance and wealth management56%B2B software and tech43%Retail and fashion28%Health, beauty and fitness19%Events and corporate retreats19%Travel and leisure18%Coaching and advisory12%Real estate8%
Month by month (Perplexity)
MonthSources per answerRankings among sourcesYouTube
July 2026 (partial month)8.437%1%
August 20269.235%0%
September 202610.434%0%
October 2026 (partial month)10.321%1%

Perplexity is built around citations, and the data shows it. Across 106,045 answers to buying questions, mostly in French, it cites at least one source in 92.3% of answers and 9.7 sources per answer, against 4.6 for ChatGPT. Its answers are also short, with a median of 217 words.

Brand sites, media and blogs make up 60.2% of Perplexity sources and rankings and comparisons 33.4%. Reviews and directories reach 5.3%, the highest share of the six engines. YouTube (0.4%), LinkedIn (0.3%), forums (0.3%) and Wikipedia (0.2%) barely register.

Rankings dominate in some sectors: 57.0% of Perplexity sources for marketing agencies, 55.6% in finance and 43.4% in B2B software, where reviews and directories also reach 8.1%. In real estate rankings drop to 8.4%.

Over the period, sources per answer rose from 8.4 in July to 10.4 in September and 10.3 in October, while the rankings share went from 36.5% to 21.3% (October covers only its first 11 days). The brand cited first changes from one week to the next in 64.2% of cases. In a head-to-head test with ChatGPT on identical questions, Perplexity cited a source in 99.1% of answers. More on our Perplexity optimization service.

Method

406,626 AI answers collected every day from July 12, 2026 to October 11, 2026 on over 1,700 purchase-intent questions tracked by Archipel (mostly in French, plus the United States and the Netherlands). ChatGPT, Google AI Overviews, Perplexity and Copilot: Profound data. Claude and Gemini: Peec AI data. Sources are classified automatically from their address and title. Figures are aggregated: no brand or client is published. Free to reuse with a link to this page.

FAQ

Frequently asked questions

How many sources does Perplexity cite per answer?
Perplexity cites 9.7 sources per answer on buying questions, the highest of the six engines tracked by the Archipel Observatory. It cites at least one source in 92.3% of answers.
Which sources does Perplexity cite most?
Perplexity cites brand sites, media and blogs for 60.2% of its sources, rankings and comparison pages for 33.4% and reviews and directories for 5.3%, based on 106,045 answers.
Does Perplexity rely on rankings and "best of" articles?
Yes, Perplexity relies on rankings and "best of" articles for 33.4% of its sources overall, and up to 57.0% for marketing agency questions and 55.6% for finance questions.
How long are Perplexity answers to buying questions?
Perplexity answers to buying questions have a median length of 217 words, shorter than ChatGPT (310 words) and Copilot (437 words), while citing more sources.

Shall we get started?

Free GEO audit: your rank on ChatGPT, Claude and Google AI Overviews, your competitors and your first actions. Reply within half a business day.

Get my free GEO auditBook a 30-min call