Profound

Profound is the answer-engine optimisation platform — the SEO suite for a world where the answer is written by a model instead of listed as ten links. Answer Engine Insights runs a tracked set of prompts against ChatGPT, Perplexity, Google AI Overviews, Gemini, Claude, Grok, Copilot, and DeepSeek and reports where you appear, in what tone, and which sources each engine cites. Prompt Volumes estimates what people actually ask them. Agent Analytics hooks into your CDN — Cloudflare, Vercel, Fastly, Akamai — to show which AI crawlers fetch which pages. The newer layer, Agents and Aim, turns the findings into work: content rewrites, technical fixes, a weekly list of the highest-impact tasks. It plugs into Search Console, GA4, the main CMSs, and, for its own retrieval, Exa and Parallel — both in our stack.

We track it rather than run it. Our own catalogue pages are built for answer engines — server-rendered links, JSON-LD, an llms.txt — and Search Console still cannot tell us whether any of it shows up inside an answer. Profound is the best-funded attempt to measure that layer end to end — a $96M Series C at a $1B valuation in February 2026, with Ramp, Figma, MongoDB, and Gamma among the customers it names. The question we are watching is whether AI visibility becomes its own budget line, the way SEO did, or folds back into the SEO suites within a year.

The direct alternatives are Peec AI and Scrunch AI on the dedicated side, and the AI visibility modules now inside Semrush — which we run — and Ahrefs, also in this Lab. Otterly.AI and Goodie cover the cheaper end. Profound is the wrong call if AI search is not yet sending you measurable traffic: Starter is $99 a month, billed yearly, for ChatGPT alone, and the three-engine tier starts at $399. It earns a look when AI answers are already sending you customers and you need to know which pages, prompts, and citations are doing it.