Vercatus vs Profound: Counting AI Answers or Changing Them

Vercatus and Profound both work on how AI answers mention brands. Compare them by the job your team needs done rather than by the length of a feature list: measuring the answers, explaining why a competitor is picked, making the change, and judging honestly whether it worked.

In short

• Compare by the job: measure, explain, change, read the result.
• Profound states a broad operating model; Vercatus is built around one loop from the answer to a measured change. The right choice depends on whether you want breadth or that loop.
• Run the same decision-focused evaluation on both, on your own topic.

This page is published by Vercatus. It is not a hands-on benchmark of Profound. Every statement about Profound’s engines, plans and price is read from Profound’s own pages, linked in the table below, and was not tested. Profound documents its own product in its help centre; check the current scope there and in a quote.

What each tool publishes, verified 12 October 2026

Each cell is read from the page linked in its row, on the date in its last column, and was not tested. Amounts are as the vendor prints them, with the billing period the page states; a cell that says a figure is not published means we found none on that page. Vendors change plans often, so check the source before you buy. This page is published by Vercatus.

Tool	Engines covered	Prompts or topics per plan	Published price	Evidence type	Source	Checked
Vercatus	ChatGPT, Google AI Overviews, Perplexity, Gemini, Claude, Grok, Le Chat, You.com, Google AI Mode and Copilot Search, every engine on every plan	Set by plan as topics, sites and polled answers a month; each plan’s allowance is listed in the plan reference	By plan and billing period on the pricing page	Answers collected on your topics with web search on; the pages engines read; AI-crawler access probes; a verdict on each shipped change	Pricing and methodology	12 October 2026
Profound	Trial: ChatGPT, Gemini and Google AI Overviews. Enterprise: up to nine answer engines. The features page also lists Perplexity, Claude, Copilot, Grok and DeepSeek	Trial: 50 prompts, run daily for 7 days. Enterprise: a tailored prompt tracking plan, count not stated	Trial: free. Enterprise: custom, no figure published	Tracked prompts (Answer Engine Insights, Prompt Volumes), AI-crawler and traffic analytics (Agent Analytics), content workflows (Agents)	Profound pricing and features	12 October 2026

A tracker counts; the question is what happens next

Most AI visibility tools start where a tracker does: a set of prompts, the answers to them, and how often a brand is named or cited. That is useful, and it is the first step of Vercatus too. The difference is what a product does once the count is in.

Vercatus is built around one loop:

1. Topics. What your market asks AI, as clusters of questions, each with an estimate of the AI demand behind it, shown as a range.
2. Answer Share. The share of AI answers on your topics that name or cite you, with its range, next to your competitors on one scale. See how Answer Share is measured.
3. Why. What each engine says without searching, the pages it reads and who they name, and the searches it runs and who ranks for them. See Why.
4. Plays. The exact change, on your site or off it, with steps, a draft to edit (or a full page written on request) and an acceptance check a machine runs. See plays.
5. Verdicts. Every shipped play is read against topics left alone and graded: Won, Lost ground, No effect, Inconclusive or Too early. See the verdict design.

Questions to ask any vendor, including us

• Population. Which answers does a share count? Two percentages called visibility are not comparable until the questions, engines and markets behind them match.
• Range. Does the number come with its range? A share read from a handful of answers moves by chance.
• Reason. When a competitor is picked, can the tool show why: the pages the engine read, the searches it ran, what it says from memory?
• Change. Does it hand you a specific change with a draft, or a general recommendation?
• Proof. After you ship, how do you know it worked? Does anything compare the topics you changed with topics you left alone?
• Governance. Seats, roles, single sign-on, data location and a DPA, in their true states. Vercatus lists its controls on the security page; SOC 2 is planned and is not represented as held.

Run a decision-focused evaluation

Pick one topic that matters and one competitor that keeps being recommended. Ask each vendor to show the answers, the reason the competitor is picked, the change they would make, and how they would tell whether it worked. Have the person who would ship the change judge whether it is specific enough to act on.

Record what each product showed, what it could not show, and the date you checked. Neither a dashboard score nor generated content on its own establishes that a change worked.

What Profound says it covers

Read from Profound’s features page on 2 October 2026 and not tested:

Area	As Profound describes it	Where Vercatus sits
Visibility	Answer Engine Insights: track brand visibility, sentiment and citations across answer engines	Answer Share with its range, engine by engine, named and cited apart
Demand	Prompt Volumes: what users ask AI	AI demand per topic, as a range with a grade
Crawler analytics	Agent Analytics: AI crawler activity and AI traffic	A site audit across nine machine identities, plus crawler traffic from your edge or logs
Content workflows	Agents: automated content workflows for AEO content	Plays with steps, a draft and, for pages on your site, a full page written on request that you edit and publish yourself; Vercatus never publishes for you
Shopping	Monitor how products appear in ChatGPT Shopping	Store topics are read like any other (product, price and review sources); no ChatGPT Shopping view is described
Proof of change	Not described on that page	A verdict on every shipped play, against topics left alone

Profound’s page lists the platforms ChatGPT, Perplexity, Claude, Gemini, Google AI Overviews, Copilot, Grok and DeepSeek, and lists enterprise controls including SOC 2 Type II and single sign-on. Vercatus reads ten engines (ChatGPT, Google AI Overviews, Perplexity, Gemini, Claude, Grok, Le Chat, You.com, Google AI Mode and Copilot Search), and its security page states each control in its true state, with SOC 2 planned and not held. “Not described on that page” means only that the page does not say so; ask the vendor.

Who each is likely to suit

Your situation	Likely better fit
You want a broad AI marketing operating model, including content workflows and shopping	Evaluate Profound’s documented scope
You want to know why a competitor is picked, get the exact change and know whether it worked	Evaluate Vercatus
Your team cannot act on recommendations	Neither alone: an agency or more capacity first
You need a particular compliance attestation today	Check each vendor’s current controls against your list

Questions people ask

Can I use both?

Yes. Some teams use a broader platform for monitoring and a focused loop for the change; check whether the overlap is worth the cost.

How should I test them?

Pick one topic that matters and one competitor that keeps being recommended. Ask each vendor to show the answers, the reason, the change and how they would read the result. Record what each showed and the date.

Is this page neutral?

No. It is published by Vercatus. It tries to be fair by quoting Profound’s own page and by saying what it did not test.

Vercatus’s pricing page states its plans, and how it works walks through the loop. For a broader shortlist, see Vercatus alternatives and the approach comparison, which explains when a tracker, an SEO suite, an agency or in-house scripts may be the better fit.