Teams almost always guess their biggest time sink wrong. Ask a finance lead where the hours go and you will hear about month-end. Sit with the team for a fortnight and you often find the real cost is somewhere much smaller and more frequent. An audit replaces the guess with a list.

Here is what happens during one, and what we usually find.

What we look at

We sit with the people doing the work, not just the people managing it. For two weeks we map every task that repeats: what starts it, which tools it touches, who is involved and where it waits. No software is installed in this phase, and nothing changes in how the team works. We are only watching and asking questions.

How we time the work

Self-reported time is unreliable, so we combine three things. We shadow a few real runs of each task, we ask people to keep a light log for a week, and we look at the timestamps in the tools themselves. Where the three agree, we trust the number. Where they don't, we ask why, and that conversation is often the most useful part of the whole audit.

What usually turns up

The details differ, but the patterns repeat:

• Copying between two tools that never talk. The same data typed into the CRM, the accounts software and a spreadsheet.

• The spreadsheet only one person understands. Critical, fragile and impossible to cover during a holiday.

• Approvals that wait for days. Usually because the approver is busy, away or doesn't know the request is waiting.

• Checking work that is almost always right. Hours spent reviewing things that have not been wrong in months.

• Chasing. Reminders, follow-ups and status questions that nobody counts as work.

The biggest wins are rarely the most complicated tasks. They are the most frequent ones.

What we don't fix

Not everything should be automated. Some tasks happen so rarely that a clear checklist beats any software. Some exist only because of an old decision nobody remembers, and the right answer is to stop doing them. Some need a conversation, not a system. Part of an honest audit is saying so.

What we fix first

We score every task on three things: hours per month, how often it goes wrong, and how risky it would be to change. The first build is the one with the most hours and the least risk. It pays for the next one, and it gives the team a working example before anything more sensitive is touched.

What you get at the end

A short report, not a slide deck. It lists every repeated task we found, the hours behind it, what we would do about it and a fixed price for each build. You can take it to us, to another team or use it to fix things yourselves. Either way, you know where the time goes.

Sample article for the Relay template. Replace it with your own writing.