Multiply the people doing a task by their weekly hours, the working weeks in a year and a fully loaded hourly cost. Then subtract the share a person will still do after automation, and weigh what is left against the cost to build and run it. In our worked example that comes to 621 hours a year and a payback of 8 months.
Count the hours honestly
Start with the people who do the task and the hours each spends on it in a typical week. Ask them, rather than estimating from a desk, and include the time spent fixing mistakes and chasing missing information.
Use 46 working weeks, not 52, to allow for leave and public holidays. Use a fully loaded hourly cost, which includes employer contributions, equipment and office costs, not salary divided by the hours in a year.
Subtract what a person will still do
Automation almost never removes all of a task. Someone still handles exceptions, approves the unusual cases and checks what the system did. Estimate that share honestly. Our calculator will not let it fall below five per cent, because a task needing no person at all afterwards is very rare.
The worked example
These inputs are illustrative, chosen to look like a typical small office task. They are not a quote and not a benchmark. Put your own numbers into the automation ROI calculator.
| Item | Value |
|---|---|
| People doing the task | 3 |
| Hours each, per week | 6 |
| Fully loaded cost per hour | GHS 60 |
| Share still done by a person | 25% |
| Hours a year today | 828 |
| Hours a year after automation | 207 |
| Hours handed back a year | 621 |
| Assumed build cost | GHS 18,000 |
| Assumed running cost a month | GHS 600 |
| Net saving a year, after running cost | GHS 30,060 |
| Payback | 8 months |
Saved hours are not always saved money
The saving is only cash if the time is used for something else that earns or saves money, or if a hire is avoided. A team that keeps the same people and simply does more has gained capacity, not cash. That can be worth more, but it should be described honestly in the business case.
If the net saving does not cover the running cost, the automation never pays back, however well it works. That is the most useful answer the calculation can give, because it stops a project before it starts rather than after.
Put a number on the task before anyone builds anything. Count real hours at real cost, keep the share a person still does, include the cost to build and run, and be honest about whether the saving is time or money.
How do I calculate the cost of a manual task?
Multiply the number of people by their weekly hours on the task, by 46 working weeks, by a fully loaded hourly cost. That is the annual cost today, before any automation.
What is a fully loaded hourly cost?
The full cost of an hour of someone's time: salary plus employer contributions, benefits, equipment and a share of office costs. It is higher than salary divided by hours worked, and using salary alone understates the saving.
Does automation pay for itself?
Only when the hours it hands back, at your real hourly cost, exceed what it costs to run, and recover the build cost within a period you are comfortable with. For low volume tasks it often does not, and the calculation shows that before money is spent.
Turn the research into a working plan.
Bring the evidence, constraints, and next decision. We will help you turn it into a clear technical brief.