Profound is the most feature-complete product reviewed in this category so far, and the least knowable. Its scope extends beyond what most competitors attempt, and it tells you nothing about what any of it costs.
What it does that others do not
The standard offering here is prompt tracking: ask engines questions on a schedule, record whether the brand appears. Profound does that across eight named engines, which is broad coverage.
Two things go further. Prompt Volumes addresses the question the rest of the category quietly avoids, which is what people actually ask AI engines. Every tool lets you define prompts to track, and almost all of them leave you guessing whether those prompts resemble real demand. Volume data, if the sample behind it is sound, changes prompt selection from guesswork into research.
Agent Analytics is the more distinctive feature. It tracks how AI crawlers fetch and interpret your site, rather than only what engines say afterwards. This matters more than it sounds, because it is the one part of AI search visibility with a clear causal chain. If ChatGPT’s crawler cannot retrieve your pricing page, no content strategy will get your pricing cited. Fixing retrieval is deterministic work in a category otherwise full of correlation.
The pricing problem
There is no pricing page. No tiers, no starting figure, no free trial that would let you form a view before speaking to someone.
For an enterprise vendor this is normal, and it is still a real cost to buyers. You cannot shortlist on budget, you cannot compare against a competitor without running two parallel sales processes, and the time spent discovering that a tool is out of range is time not spent on the work. In a category this young, where nobody has established what the software is worth, opacity favours the seller.
The practical response is to ask for a number on the first call and treat hesitation as information. AthenaHQ’s published 295 dollars a month is currently the only public anchor in this category, which makes it useful as a benchmark even if you have no intention of buying it.
What to weigh
The autonomous agents framing is the part to discount. Describing automation as autonomous workers for every function of a marketing team promises a degree of independence that nothing in this category currently justifies, and the honest version of these products is measurement with suggestions attached.
Only one named customer reference appears publicly, so breadth of adoption is difficult to assess from outside. That is not evidence against the product; it is a limit on what a documentation review can establish, and worth naming.
If you are an enterprise with the budget already approved and the crawler-level data appeals, this belongs on the shortlist. If you need to know the price to decide whether to shortlist it, start elsewhere and come back.
