Loading
At this level the answer set is the product. Writing the top twenty questions properly usually pays for itself in a quarter.
Distinct questions is the input people guess worst. Take a week of tickets and group them honestly; it is almost always a smaller number than expected.
It identifies the questions your team answers repeatedly — internally or from customers — and costs them out over a year.
Individually each is trivial, which is why nobody addresses them. Annually they are usually somebody's full-time job.
Sample two weeks of questions from wherever they arrive: support tickets, internal channels, email. Group them by what is actually being asked rather than by wording.
Estimate frequency and the time each takes, including the interruption cost to whoever answers.
It will not tell you the answer should be automated. Some repeat questions are a signal that something is confusing and should be fixed at source.
It also cannot see questions people stopped asking because they gave up.
Sort by total annual time rather than by frequency. A question asked weekly that takes twenty minutes outranks a daily one that takes thirty seconds.
The top few are your documentation backlog, and they are also the questions an assistant could answer once the article exists.
Two weeks is usually enough for the pattern.
Total time. It reorders the list.
No.
Half an hour on your own figures, and an honest answer about the parts Treepie does not improve.