Writing
Which repeated work is worth removing first
Not every dull task is worth automating. The right one to start with has three marks, and being the most tedious is not one of them.
Published Revised One Other
Where do you start?
With the task that happens often, happens the same way every time, and shows its mistakes immediately. The three only count together. Without the third you save time and lose trust.
Copying data from one place to another meets all three. Handling a difficult client meets none.
List what you did more than three times this week. The right task is usually on it, and rarely the one that came to mind first.
Do you need AI for this?
Usually not. A spreadsheet wired to a form removes more repeated work than any model, and it fails in ways you can understand.
AI earns its place where the job is reading text and deciding what to do with it. Sorting messages, summaries, first drafts. Work where an approximate answer is useful and you read it before it leaves the building.
I say that with the machine beside me all day. It is good at reading and drafting, and the judgement about what goes out stays with a person.
What if the tool breaks?
It will. The useful question is not whether it breaks, but what breaks with it and how long you take to notice.
Automate first what you can check at a glance and undo by hand. Leave money, contracts, and anything that goes straight to a client for later.
How do you know it was worth it?
Measure before. One week counting how often you do the task and how long it takes is enough to know what you are saving.
If nobody has gone back to the old way after a month, it stuck. If they have, time was not the problem. Something else was, and finding it beats defending the tool.
An automation nobody uses is not neutral. It stays there, broken, and nobody says so. Switching off what did not take hold is part of the job.
Send me an email with the task and I will tell you whether it is worth removing.