Most audits produce a list. A list is easy to write and difficult to act on, because there is no way to tell whether anything is missing. Three lenses solve that. Every finding maps to one of them, and a lens with nothing under it is a question.
Time lost is the hours consumed by work no customer would pay for. Status meetings. Reports assembled by hand from four systems. Chasing an update that already exists in writing somewhere. It is the easiest lens to price, because people can usually give you the hours once the question is precise.
Work repeated is the same output produced more than once, because the first version could not be found or could not be trusted. Client onboarding written from scratch each time. An answer given again to a question answered last month by someone who has since left. It is the lens that scales worst: it grows with headcount, which is why it usually surfaces as a ceiling on growth and gets recorded nowhere as a cost.
Waiting on information is decisions that stall because the number, the file or the approval is elsewhere. It is the lens most often left unpriced, and it is usually the largest, because nobody counts the hours a decision spent waiting. The clearest symptom is a single sentence we hear in almost every audit: getting a specific number out of the business takes about an hour. This is also the lens that most shapes how a Cortex audit is run, because it is the one nobody volunteers.
The lenses matter because they change what gets built. A finding under time lost usually becomes a visibility system. A finding under work repeated usually becomes a knowledge system. A finding under waiting on information usually becomes a connection between two things that already exist and have never spoken to each other. Those are three of the five kinds of system Cortex builds, and the lens decides which one a company needs.
And a lens with nothing under it means the interviews missed something, which is a more useful result than a longer list.