Three questions to ask before automating anything
Before you build a single integration, three questions will tell you whether the automation is worth it — or whether you should delete the work instead.
Enthusiasm for automation tends to skip the most important step: deciding whether the work should be automated at all. Three questions, asked before any building, will save you from automating the wrong thing well.
1. Should this work exist?
The fastest automation is deletion. Plenty of tasks are vestigial — a report no one reads, a check that no longer matters. Before automating, ask whether you could simply stop doing it. If yes, you're done.
2. Is the process stable?
Automating a process that changes every week means rebuilding every week. If the workflow is still in flux, stabilize it first. Automate the version you'll actually keep.
3. Does the math work?
Frequency times time times people, against the cost to build and maintain. High-frequency, rules-based work clears this bar easily. A rare, complex task usually doesn't.
Delete first, stabilize second, automate third. Most teams jump straight to third and pay for it.