Surfais and Profound compared on engine access, price, security and published method, with figures read from both vendors' own pages on 6 September 2026.
Verified
6 September 2026
| Dimension | Surfais | Profound |
|---|---|---|
| AI engines included | All five on every plan, including the £69 Starter: ChatGPT, Claude, Gemini, Perplexity and Google AI Overviews. | Starter is ChatGPT only. Growth adds Perplexity and Google AI Overviews, for three. Enterprise goes up to nine, including Gemini, Claude, Copilot, Grok and DeepSeek. |
| Engine gating | None. No engine add-ons and no engine unlocks by tier. | Yes, and it is the steepest in the category: 1 engine, then 3, then up to 9 by tier. Claude and Gemini require Enterprise, which is custom-priced. |
| Entry price | £69/mo Starter, ex-VAT. Growth £199, Scale £479, Enterprise from £1,999. 15% off annual. | $99/mo Starter. Growth $399, Enterprise custom. Annual billing gives two months free. |
| Prompts at entry price | 20 prompts, 1 brand, 1 country, 2 competitors. Extra prompts £6 each on brand tiers. | 50 prompts and 1,500 responses at Starter, 1 language, 1 region. |
| Seats | Not published. Plans are scoped by brands, prompts, countries and competitors rather than by seat. | 1 seat at Starter, 3 at Growth. |
| Refresh cadence | Twice weekly on brand tiers, weekly on Agency Mode, daily on Enterprise. | Daily. |
| Published score methodology | AIS Score weights are published and versioned. Version 2, dated 10 June 2026: Visibility 60%, Position 15%, Sentiment 10%, Share of voice 15%. Published at surfais.com/methodology. | Profound publishes a method for the free public Profound Index research product, covering 400M+ conversations, entity extraction models and weekly recalculation. No weighting or versioning is disclosed for the paid-product score, and tryprofound.com/methodology returned a 404 on 6 September 2026. |
| Free trial | 7 days, no card. 1 brand, one weekly scan plus two manual scans, one audit. | A Try for free call to action exists on Growth. Length and card requirement are not published. |
| Agency / white-label | Agency Mode at £59 per client brand, minimum 5 brands, 20 prompts each, pooled. White-label reporting is scheduled to ship at the end of September 2026 and is not live. | An Agency Growth variant is referenced at $399/mo with 400 credits per month per client workspace. No white-label programme found. |
| API and integrations | No public API, no MCP server and no Looker Studio connector. | API at Enterprise only, alongside SSO/SAML, a dedicated Slack channel and a 24-hour SLA. No API on Starter. |
| Security certifications | None. No SOC 2 and no ISO 27001 — SOC 2 is budgeted, not achieved. No SSO or SAML. | SOC 2 Type I confirmed, with a public trust centre at trust.tryprofound.com. We could not verify Type II. |
| Company stage | Early-stage UK company. Small customer base, no published customer count, no named enterprise logos. | $96M Series C on 24 February 2026 at a $1B valuation, led by Lightspeed with Sequoia and Kleiner Perkins. Publishes an enterprise customer list. San Francisco. |
Verified 6 September 2026. Prices and features in this category change often; check each vendor's own site before you buy. Prices are shown in each vendor's own published currency and are not converted.
Profound's engine gating is the steepest in this category, and it is the row that decides most comparisons. At $99 you are tracking ChatGPT alone. At $399 you are tracking three engines. Claude and Gemini arrive only at Enterprise, which is custom-priced and therefore a sales conversation rather than a checkout. Surfais includes all five — ChatGPT, Claude, Gemini, Perplexity and Google AI Overviews — at £69, and the list does not change as you move up the tiers.
For a brand whose buyers ask Claude or Gemini as often as they ask ChatGPT, that is not a pricing nuance. It is the difference between a measurement and a sample of one.
The second difference is disclosure of the score. Profound publishes a genuine methodology for its free public Index — a serious research product built on more than 400 million conversations, recalculated weekly. The paid product's score is a different matter: no weighting is disclosed, no version is published, and the methodology URL returned a 404 when we checked on 6 September 2026. Surfais publishes the AIS weights, versions them, and recalibrated them downward in June 2026 against its own commercial interest after finding a customer with six mentions in sixty-nine runs was being scored as competitive.
Profound is the best-capitalised company in this comparison and it has built accordingly.
Security. SOC 2 Type I is confirmed and there is a public trust centre. Surfais has neither SOC 2 nor ISO 27001 — SOC 2 is budgeted out of our round, not achieved — and no SSO or SAML at all. If your security review has a hard gate, Surfais does not clear it and Profound does.
Engine breadth at the top. Enterprise reaches up to nine engines including Grok and DeepSeek. Surfais covers five and does not go further at any price.
Cadence. Daily against our twice weekly below Enterprise.
Beyond measurement. Profound has moved toward an action layer — marketing agents intended to change what engines say, not only report it. Surfais measures and audits; it does not act.
Standing. A $96M Series C in February 2026 at a $1B valuation, led by Lightspeed with Sequoia and Kleiner Perkins, and a published enterprise customer list. Add an API, SSO, a dedicated Slack channel and a 24-hour SLA at Enterprise, and a free public benchmarking index that is genuinely useful whether or not you buy anything. Surfais is an early-stage company with a fraction of the customers and none of that enterprise apparatus.
Choose Profound if you are an enterprise buyer with a procurement process. If SOC 2 is a gate, if SSO is non-negotiable, if you need an API and an SLA, or if you want an agent layer that tries to move the numbers rather than only report them, Profound is the better fit and Surfais will not pass your review.
Choose Surfais if you are a brand or agency that wants all five major engines tracked from the first month at a price you can approve without a committee, and a score whose formula you can read. At Profound's $399 Growth tier you are tracking three engines; at £199 Surfais tracks five across two brands and two countries.
Put plainly: Profound is built for the buyer who needs the paperwork. Surfais is built for the buyer who needs the coverage.
Profound, based in San Francisco, has shifted its positioning toward marketing agents for winning in AI search. It is no longer only a measurement tool; it aims to be an action layer that produces and tunes content in response to what engines say. Behind it sits a $96M Series C raised in February 2026 at a $1B valuation, a SOC 2 Type I attestation, a public trust centre and a published enterprise customer list.
Surfais is a UK measurement product. It tracks five engines, scores what they say on a published scale, and produces a monthly deep-dive audit on every plan. It does not generate content and it does not act on your behalf.
Profound's tiering is unusually explicit about what you are buying: one engine, then three, then up to nine. There is a defensible logic to it — running prompts against nine engines daily is genuinely expensive, and charging for that cost is not unreasonable. But it has a consequence for the buyer that is easy to miss during a trial. A Starter customer measuring ChatGPT alone will form a view of their AI visibility that may not survive contact with Claude or Gemini, and nothing in the product will tell them so.
Surfais took the opposite decision: five engines on every plan, and the rationing happens on prompt volume instead. Twenty prompts at £69 is a premium per prompt, and we would rather defend that than pretend otherwise. The argument is that twenty well-designed prompts measured across five engines tell you more than fifty prompts measured against one.
It would be unfair to say Profound does not publish method. They publish a substantial one for the Profound Index, their free public research product: more than 400 million conversations, entity extraction models, weekly recalculation. That is more open research than most companies in this category fund at all, and it is available whether or not you are a customer.
The narrower and more accurate point is about the paid product's headline score. We found no disclosed weighting, no version history and no reproducible definition, and tryprofound.com/methodology returned a 404 on 6 September 2026. The IAB's August 2026 framework describes measurement as decision-grade when it is reproducible, disclosed and versioned. On that specific test — the headline score, not transparency in general — Surfais meets it and, as at the date of checking, Profound's paid score does not.
These products are not really competing for the same signature. Profound is selling to an enterprise marketing function with a security review, a procurement cycle and an appetite for an agent layer. Surfais is selling to brands and agencies who want complete engine coverage early, at a price that does not require a business case, and who want to be able to ask how a number was calculated and get an answer with a version number on it.
If both of those descriptions fit you, take the security row seriously: no SOC 2 and no SSO is a real limitation, not a footnote, and it is the honest reason a large enterprise should choose Profound today.
Verified 6 September 2026. Prices and features change; check the vendor's own site.