Operations and planning
Spotting recurring tasks
Most time savings lie in tasks that come back every week. How to spot them and choose where AI or automation pays off most.
5 min read
AI pays off most on tasks that come round often. But which are they exactly? In the daily rush, you stop seeing them. A short exercise brings them to light.
Log a week
For one week, note what you do, per half hour or per task. It doesn't need to be a perfect log; keywords are enough. Ask colleagues to do the same if you want to look at the team.
Have it analysed
Paste your list and ask the AI to:
- Group similar tasks.
- Estimate how much time each group takes per week.
- Indicate how often each group recurs and how predictable it is.
Choosing where to start
For each group, ask three questions:
- How often? Daily or weekly pays off more than monthly.
- How predictable? Always the same steps suits automation. Different every time suits an assistant or a good prompt.
- How annoying? Tasks nobody likes are a good start. You already have support.
From task to approach
- Text-heavy tasks that differ a little each time: a standard prompt or an assistant.
- Tasks with always the same steps between programs: an automation.
- Tasks that mainly require decisions: keep them with people, but let AI do the groundwork.
Exercise
Your answers stay only in this browser.
Prompt to copy
Below is what I did over one week. 1) Group similar tasks. 2) Estimate the time per week for each group. 3) Indicate how often each group recurs and how predictable it is. 4) Suggest the three groups that would gain most from a standard prompt, an assistant or an automation, and why.
[paste your list]
