Manual data entry is easy to notice and hard to price accurately. Multiplying minutes by salary is a useful start, but it misses rework, waiting time and the value of work that never gets done.
Build a defensible baseline
Sample the workflow over several normal weeks. Record volume, active handling time and the proportion that requires correction or a follow-up.
Use a fully loaded hourly cost where available, but keep assumptions visible. A range is more honest than a precise figure built on weak evidence.
- Transactions per week
- Minutes of active handling per transaction
- Error and rework rate
- Elapsed delay created by the handoff
- People and systems involved
Include the cost of interruption
A two-minute update may take longer in practice when it interrupts focused work or waits in somebody’s inbox. Record how the task arrives and whether work must be reconstructed before it can continue.
Opportunity cost should be described carefully. Do not count every saved minute as new revenue; identify what the capacity could realistically be used for.
Compare against the whole change cost
Implementation, licences, maintenance, training and exception handling all belong in the comparison. A cheap-looking automation can be expensive if it is fragile or depends on constant correction.
Start with a small trial and confirm the assumptions. The purpose of the calculation is a better decision, not a larger headline saving.