All tools

Tool · Decide

Should you actually automate this?

10 questions, 2 minutes. Real go / no-go logic — not a vendor pitch. Built from 120+ automation engagements where we wish someone had asked these first.

1
Is the process rule-based with low judgment required?

Predictable, deterministic paths automate cleanly.

2
Is the volume high enough to justify build cost (>5k/year)?

Low volume = manual stays cheaper.

3
Is the input data structured and digitally accessible?

PDFs and emails need a parsing layer first.

4
Has the process been stable for 6+ months?

Automating a moving target is a sunk-cost trap.

5
Is there a single accountable process owner?

No owner = no one to approve exceptions.

6
Is the current process documented end-to-end?

If not, you're automating tribal knowledge.

7
Are exceptions <15% of total volume?

Exception logic dominates build cost.

8
Do the underlying systems expose APIs (not just screens)?

API > RPA scraping for stability.

9
Is the regulatory / audit complexity manageable?

Heavy compliance needs human-in-the-loop design.

10
Will saved hours go to higher-value work (not just headcount cuts)?

Pure cost plays usually disappoint.