Updated
Updated What changed on
Fact-checked Every UK share figure, price and crawler claim re-checked on StatCounter, OpenAI and Perplexity's own pages on 14 October 2026, with prices confirmed on a second official page. Study figures re-read on Profound, Ahrefs and Search Engine Journal, and the newer Reddit figures and Perplexity's retrieval description re-checked on Ahrefs' and Perplexity's own pages in the same 14 October 2026 source checks.
- UK share now uses StatCounter's September 2026 referral figures: ChatGPT 80.02%, Gemini 10.14%, Perplexity 2.7%, replacing December 2025 figures StatCounter's data no longer reproduces.
- ChatGPT pricing now shows UK prices: Go £7, Plus £20 and Pro from £89 a month, with ads on Free and Go since the UK launch.
- Perplexity pricing now includes Max at $200 a month and notes that its listed prices are in US dollars, with local currency offered at checkout.
- Reddit figures corrected: Profound puts Reddit at 6.6% of Perplexity citations and 1.8% of ChatGPT's, replacing a misread 24-46.7% range, and the source-mix chart now uses those numbers.
- Crawler advice corrected: OAI-SearchBot, not GPTBot, decides ChatGPT search visibility, and the unsupported Google paid tier claim is gone.
- Agentic section updated: ChatGPT agent is gone and OpenAI points users to ChatGPT Work, Comet is free to download, and agentic booking reached UK AI Mode in April 2026.
- Reddit picture brought up to date: in Ahrefs' September 2026 US data reddit.com is the top ChatGPT domain at 16.8%, with Wikipedia second at 7.0%, and a separate Ahrefs US dataset showed ChatGPT's Reddit citations falling 88% from March to September 2026. The Profound figures are now labelled as 2024-25 data.
- Perplexity's retrieval now reads as its own web index combined with partner search APIs, as Perplexity describes it.
ChatGPT vs Perplexity: which to optimise for first.
The 30-second verdict
For UK B2B audiences, ChatGPT is priority one (about 80% of UK chatbot referrals in September 2026 against Perplexity's 2.7%, per StatCounter). For research-heavy, technical or academic audiences, Perplexity's footnote-citation transparency makes it worth more than its share suggests. By my estimate about 70% of the optimisation work overlaps. The 30% that differs: OAI-SearchBot access and Bing visibility for ChatGPT, accurate dates + named-author signal for Perplexity. Most UK B2B brands should optimise both with weighted attention.
Side-by-side comparison
 ChatGPT |  Perplexity | |
|---|---|---|
| UK chatbot referral share (StatCounter, Sept 2026) | 80.02% (CSV export: 79.8%) | 2.7% (CSV export: 2.74%), 4th behind Gemini and Copilot |
| Audience skew | Mainstream consumer + B2B | Research-heavy / technical / academic |
| Citation prominence | Partial - web search answers "may include citations" (OpenAI) | Always - "every answer includes inline citations" (Perplexity) |
| Reddit share of citations (Profound, Aug 2024 to June 2025) | 1.8%. Ahrefs found pages from its Reddit source cited only 1.93% of the time (reported by SEJ) | 6.6% (Reddit is Perplexity's top cited source) |
| Live web crawler | OAI-SearchBot + third-party search partners (OpenAI names Microsoft) | PerplexityBot + its own web index, combined with partner search APIs |
| Recency weighting | Orders references newest first | Orders references newest first, but cited pages still average 1,166 days since publication (Ahrefs measured 17 million citations, July 2025) |
| Top cited source (Profound, Aug 2024 to June 2025) | Wikipedia, 7.8% of all citations. In Ahrefs' September 2026 US data, reddit.com was first at 16.8% of mentions, with Wikipedia second at 7.0% | Reddit, 6.6% of all citations |
| Pricing for users (checked Oct 2026) | UK: Free, Go £7, Plus £20, Pro from £89 a month. Free and Go may show ads; Plus and above are ad-free | Free, Pro $20 a month ($200 a year), Max $200 a month. Priced in US dollars; you can pick local currency at checkout |
| UK B2B priority | 1st | 3rd-4th depending on audience |
| Optimisation lever | OAI-SearchBot access + Bing visibility + extractable content | PerplexityBot access + named author + credible sources |
When to pick which
Optimise for ChatGPT first if
Your audience is mainstream UK B2B - law firms, accountants, manufacturers, estate agents, mid-market SaaS. ChatGPT's UK share is large enough that visibility here drives more measurable conversation than any other AI engine. Use the how to rank in ChatGPT guide for the engine-specific work once the shared GEO basics are in place.
Optimise for Perplexity first if
Your audience is research-heavy - academic, technical, financial analyst, journalism, deep-dive consultancy buyers. Perplexity says every answer includes inline citations, and the audience self-selects for "I want sources I can verify".
Optimise for both (most brands)
By my estimate about 70% of the work overlaps - schema, named author, content depth, ranking foundations. Build that foundation once and you cover both engines. Then add the 30% engine-specific layer: OAI-SearchBot access + Bing visibility for ChatGPT, and PerplexityBot access + an accurate dateModified for Perplexity so a genuinely revised page is read as revised. OpenAI's crawler docs say OAI-SearchBot controls whether you appear in ChatGPT search, while GPTBot is its training crawler, so you can allow one and block the other.
The honest read on differences
The biggest practical difference is what each engine treats as a credible source. ChatGPT retrieves pages from its separate Reddit source often but rarely cites them: in Ahrefs' data, 67.8% of the pages ChatGPT retrieved but did not cite came from that source. That study ran on an older ChatGPT model, and Reddit pages also reach ChatGPT through normal search. In Profound's 2024-25 data Wikipedia was ChatGPT's top cited source and Reddit led for Perplexity. In Ahrefs' September 2026 US data, reddit.com was the top ChatGPT domain at 16.8% of mentions, with Wikipedia second at 7.0%. Perplexity also shows its citation pattern openly. If you cannot diagnose why a competitor is being cited in Perplexity, you usually can - the footnotes are right there. ChatGPT is harder to audit because the citations are less consistent.
The second difference people point to is recency, and the data is weaker than the claims. Ahrefs (July 2025, 17 million citations) found both engines order in-text references newest first, yet the pages Perplexity cites average 1,166 days since publication. Freshness seems to affect where you sit in the answer more than whether you are in it. Keep your dates accurate, but do not expect a date change on its own to win citations.
What I see when I run the same prompt through both
To make the difference concrete, three queries I ran through ChatGPT (browsing on) and Perplexity Pro on the same day. These results were published in May 2026 and were not re-run for the October 2026 update. The patterns hold across most categories I have tested.
- "Best UK GEO consultant" - ChatGPT named 4 sources, mostly agency landing pages without inline citations. Perplexity listed 6 sources with footnotes, including 2 independent consultants and a Search Engine Land article. Perplexity's response was harder to skim but easier to verify.
- "What does generative engine optimisation cost in the UK" - ChatGPT gave a paragraph answer with £-band estimates and no source attribution. Perplexity gave a structured answer citing 4 distinct pricing sources including a 3chillies blog post and a UK agency listicle. For buyer-research queries Perplexity wins on trust.
- "How to fix robots.txt for ChatGPT" - ChatGPT gave the most useful technical answer (it knows its own bot policies best). Perplexity gave a more cautious answer citing OpenAI's docs and one Stack Overflow thread. For ChatGPT-specific technical queries, ChatGPT is better.
Takeaway: prompt type drives engine preference. Buyer-research queries favour Perplexity. Technical / implementation queries favour ChatGPT. Most B2B brands need to be cited in both because buyers use both modes before they decide.
Agentic actions: where both engines are heading
Both engines are moving toward agentic flows (booking, buying, scheduling, ordering on the user's behalf) in 2026, and the products keep changing. OpenAI folded Operator into ChatGPT agent, then withdrew ChatGPT agent and now points users to ChatGPT Work for multi-step tasks, and has retired its Atlas browser. Perplexity's Comet browser has been free to download for everyone since October 2025, though the Comet Assistant agent needs a paid plan. Google AI Mode reached the UK in July 2025, and agentic restaurant booking followed in UK AI Mode in April 2026.
The practical implication for SEO/GEO: optimise the booking and transaction flows now, not just the content. Schema for ProfessionalService availability, OpeningHoursSpecification, Reservation, Order - these become directly actionable in agentic flows. Pages that an AI agent can transact against will be cited and recommended more. Pages that require human form-filling will get cut out of the agentic path.
More on the comparison.
Which is better for SEO, ChatGPT or Perplexity?
For UK B2B audiences, ChatGPT is higher priority: about 80% of UK chatbot referrals against Perplexity's 2.7% in September 2026, per StatCounter. For research-heavy, academic or technical audiences, Perplexity's footnote-style citations make it the stronger surface. Most brands need to optimise for both with different content tactics per engine.
Does ChatGPT or Perplexity cite Reddit more?
In older data, Perplexity cited Reddit more. In Profound's data (August 2024 to June 2025), Reddit made up 6.6% of all Perplexity citations and 1.8% of ChatGPT's. Ahrefs, as reported by Search Engine Journal, found ChatGPT cited pages from its separate Reddit source only 1.93% of the time, and 67.8% of the pages it retrieved but did not cite came from that source. Newer data is mixed: Reddit was the most-mentioned domain in ChatGPT answers in Ahrefs' September 2026 US data, at 16.8%, though a separate Ahrefs US dataset showed ChatGPT's Reddit citations falling 88% between March and September 2026.
Which has better citation transparency?
Perplexity. Perplexity says every answer includes inline citations. OpenAI says ChatGPT answers that use web search may include citations, so they show up less consistently. If you are diagnosing why your brand appears or doesn't, Perplexity is easier to audit.
Which has the bigger UK audience?
ChatGPT by a wide margin. StatCounter's UK chatbot referral share for September 2026: ChatGPT 80.02%, Gemini 10.14%, Copilot 5.61%, Perplexity 2.7%. Its CSV export for the same month gives 79.8%, 10.21%, 5.68% and 2.74%. It measures referral traffic, not users. Perplexity is now fourth in the UK on this measure, behind Gemini and Copilot.
Do they share content optimisation work?
By my estimate about 70% of the work overlaps - schema, named author, content depth, ranking foundations. The 30% that differs: for ChatGPT, allow OAI-SearchBot, the crawler OpenAI uses for ChatGPT search, and keep Bing visibility healthy, as OpenAI names Microsoft among its search partners; recency matters more for where Perplexity places a citation than for whether you get one. Ahrefs (July 2025) found Perplexity and ChatGPT order in-text references newest first, while the pages Perplexity cites average 1,166 days since publication.
Jason Burns
Independent UK SEO, GEO and AI consultant. In search since 2007. Portfolio includes 3M, BlackRock, Unilever and E.ON. Owner of SEO Moves Ltd since 2014.
More about Jason →Want to see where your brand sits on both engines?
Send three buyer queries plus your top competitor name. I run them through ChatGPT AND Perplexity, send the side-by-side citation diff back, then we talk about which engine deserves the bigger share of your GEO budget.
Request the dual-engine baseline