Most automation projects start with a tool. Someone sees a demo, imagines the hours it could save, and the project begins with configuration. Weeks later, the automation works — but it reproduces every unnecessary step, workaround and exception of the original process.
A better starting point is a plain description of how the work is actually done today.
Map the process as it runs, not as it is documented
Written procedures describe how work is supposed to happen. The people doing it know how it really happens: the spreadsheet that bridges two systems, the colleague who is copied on every email, the approval that is always given.
Sit with them. Follow a few real items from start to finish and write down every step, handover and wait.
Look for three things
Once the steps are visible, three questions do most of the work:
- Which steps exist only because two systems are not connected?
- Which steps add no value at all and could simply be removed?
- Which steps are rules-based and repeated often enough to automate?
Then decide what to automate
Remove first, then simplify, then automate what remains. The automation that follows is smaller, cheaper to build and much easier to maintain, because it is built around a process that already makes sense.






