Recomma vs Profound (2026): self-serve tracking vs an enterprise contract

Profound and Recomma both ask AI assistants the questions your buyers ask, mostly in the interface a buyer reads, and count how often your brand is named. They are built for different buyers. Profound is an enterprise platform sold by contract: nine engines, a dataset of what people really ask AI, crawler analytics from your CDN, content agents, SOC 2 and SSO. Recomma is self-serve from $30 a month, with fewer engines, and spends its effort on whether a figure can be trusted and whether a fix moved it.

At a glance

Profound's column is from its pricing page and documentation; Recomma's is from the plan table the product enforces.

Plans and prices

Profound's pricing page lists two plans for brands. The Trial is free for seven days: fifty prompts chosen for your industry, which you cannot edit, run daily on ChatGPT, Gemini and Google AI Overviews, with unlimited seats and limited agent credits. Enterprise is "Custom": a tailored set of prompts you can edit, up to nine engines, daily tracking, custom languages and regions, CSV and JSON exports, API access, ChatGPT Shopping, SSO/SAML, SOC 2 and a dedicated specialist with a 24-hour SLA (Profound pricing). No figure is published for Enterprise.

For agencies, the same page mentions a self-serve Agency Growth plan with 400 agent credits a month per client workspace, without a price. Profound launched it in November 2025 with client workspaces at $399 a month each, every one tracking 100 prompts on ChatGPT, Perplexity and AI Overviews, plus ten pitch workspaces a month and five seats (Profound blog). Whether $399 is still the price today is not published.

Recomma sells three self-serve plans and an Enterprise tier quoted by a person. Starter is $30 a month for 50 prompts, sampled weekly, on one project. Growth is $299 for 300 prompts, sampled daily, across three projects. Scale is $999 for 500 prompts, daily, with no limit on projects or engines. Prompts pool across the workspace rather than being allotted per project, so an agency decides which client gets the coverage instead of buying it client by client. The full table is on the pricing page.

Engines, and how answers are collected

Profound's Enterprise plan covers up to nine engines: ChatGPT, Perplexity, Google AI Mode, Gemini, Microsoft Copilot, DeepSeek, Claude, Google AI Overviews and Exa Search. Its own summary of the product, updated in September 2026, also names Grok (official information page). The same page says Profound "captures responses directly from the consumer browsing experience rather than from model APIs."

Recomma captures ChatGPT, Gemini, AI Overviews and Google AI Mode from the interface a reader sees too, and asks Perplexity through its own API. On collection method the two largely agree, and for a good reason: in our own study, only 8% of the pages ChatGPT cited in its interface also appeared in the API answer to the same question (ChatGPT's interface vs its API). Where they differ is coverage. Copilot is not sampled by Recomma, and Claude is not sampled at all.

How each turns answers into a number

Profound documents its visibility score openly: for each model, the runs that name your brand divided by the runs that name any brand, with runs that name no brand left out of both; then an equal-weight average across models, so "a model with many runs counts the same as a model with few" (how metrics are calculated). Two things follow. Profound's percentage is a share of answers that mention brands at all, so it should not be set beside a figure from a tool that divides by every answer. And the page publishes no sample size or error range to go with it.

Recomma prints the number of answers behind every figure and a 95% Wilson interval beside it, and dims a reading too thin to mean anything rather than drawing it as a trend. An AI answer is sampled, not looked up — ask twice and the brands can change — so a movement from 12% to 15% on a few dozen answers is often noise. Why the error bar and what sampling can see set out the reasoning.

Beyond tracking

Profound has grown into a marketing platform. Its main additions, each from its own documentation:

• Prompt Volumes, which Profound also calls Conversation Explorer: a dataset built by "licensing AI answer engine conversations" from ChatGPT, Gemini and Perplexity, cleaned and modelled into volumes for topics and keywords (docs; product page).
• Agent Analytics: AI crawler and bot traffic read from your CDN or server logs, with crawler classification and page-level access (docs).
• Content agents paid for in credits, ChatGPT Shopping reports, and since October an SEO tab built with Moz and a Projects workspace (changelog).

Recomma's additions are narrower, and about acting on the number:

• Each tracked question is priced with keyword data — monthly searches, the share of clicks a top result takes, and what advertisers bid — so you can see what the market behind a gap is worth.
• A gap becomes an action. When the action closes, the reading on its prompts is frozen, and fourteen days later the same prompts are asked again on fresh answers. The verdict is kept whether or not anything moved.
• The brand's own site is audited against fifteen markup rules that AI crawlers trip on, with each failing page's markup as it is and as it should be. The checklist lists all fifteen.
• Visits that assistants send to the site are reported beside the answers that sent them.

API, MCP and exports

Profound's API, CSV and JSON exports are on Enterprise. Its MCP server is too, and it reads and writes: an agent can create, edit, disable or delete prompts, start agents and publish workflows, add to knowledge bases and manage projects. The documentation does not describe a preview or confirmation step before a write takes effect (Profound MCP overview).

Recomma's read-only REST API comes with Scale; CSV export comes with every plan. Its MCP server, also on every plan, reads and writes: an assistant can list the gaps, read the site findings with their before-and-after markup, propose new prompts and close an action. Every write previews first, needs a separate write scope, and new prompts land as suggestions that spend nothing until a person accepts them. Fixing what AI search gets wrong, from Claude Code walks through it.

Where Profound is the better choice

• Engine breadth. Nine engines on Enterprise, including Copilot, Perplexity, Claude and DeepSeek. Recomma tracks five, and three below Scale.
• Procurement. SOC 2 Type 2, SSO/SAML, SCIM provisioning and a dedicated specialist with an SLA. Recomma does not publish any of these.
• What people actually ask. Prompt Volumes is modelled on licensed AI conversations; Recomma's demand figures come from search keyword data, which is a proxy.
• Crawler data. Agent Analytics reads AI bot traffic from your CDN; Recomma has nothing equivalent.
• Breadth beyond visibility: content agents, shopping reports, SEO data and documents in one platform.
• Track record. Profound says it is used by more than 700 enterprise brands and large agencies; Recomma is new and has no reviews yet.

Where Recomma is the better choice

• A price on the page. $30 a month to start, on a card, without a sales call; Profound publishes no price for a brand plan.
• Numbers you can defend. A sample size and confidence interval on every figure, so a change that is noise is shown as noise.
• Proof that a fix worked. Actions re-measured fourteen days after they close, against a frozen baseline.
• Prompts that pool. One allowance across every project, rather than a fixed workspace bought per client.
• An assistant that acts safely. An MCP server that writes behind a preview and a separate scope, on a self-serve plan.
• The site itself. A fifteen-rule audit of the markup AI crawlers read, with the fix for each failing page.

Choose Profound if… choose Recomma if…

Choose Profound if you buy through procurement and need SOC 2 and SSO on day one; your buyers use Copilot or Claude; you want to know what people ask AI, not only what search engines record; or you want crawler analytics, content agents and visibility in one contract.

Choose Recomma if ChatGPT, Gemini, AI Overviews and Google AI Mode and Perplexity cover the engines your buyers use; you want to start this week at a price you can read; you have to show a client or a board which changes moved the number, with the error bar to back it; or you work from an AI assistant and want it to fix the site behind a preview.

Other self-serve options are in Profound alternatives. The same comparison against other tools is in Recomma vs Peec AI and Recomma vs Otterly.AI, and the whole field in the best AI visibility tools.

How we compared

Profound's plans, engines, features and methods were read from its public pricing page, documentation, changelog and blog on 14 October 2026. Where Profound does not publish something — the Enterprise price, the current price of Agency Growth, a sample size or error range on its figures — the page says so rather than guessing. Its SOC 2 Type 2 is stated in its June 2025 announcement and on its pricing page. Third-party review scores were left out because they could not be checked at the source. Recomma's figures are drawn from the same plan table its API enforces, so they change on this page when they change in the product.