I review the product against usability principles and against your own analytics, then triage the findings by severity and effort. You get a prioritised list that goes straight into a backlog, not a slide deck.
You know something is wrong and can't name it. Support tickets cluster somewhere nobody expected. A funnel drops sharply at one step. Or a new team has inherited a product and needs an honest map of its problems before planning anything.
An audit is only useful if it produces a ranked list, not a catalogue. Anyone can find a hundred issues in a mature product. The work is deciding which twelve matter.
I go through the product against the jobs users are actually trying to finish, not a generic heuristic checklist — though the heuristics get applied where they earn their place. Then I pull whatever data exists: analytics, support tickets, session recordings, sales objections. Qualitative findings that agree with quantitative ones are the strongest thing an audit can give you.
Output is a prioritised list, each item with the evidence behind it and a rough sense of effort, so it can be argued with rather than simply accepted.
Two to three weeks for a product of moderate size. Faster if the scope is a single flow.
Do you need access to our analytics?
Not strictly, but the audit is substantially better with it. Behaviour data turns opinions into findings.
Will you tell us things we won't like?
That's most of the value. A diplomatic audit is a wasted one.