What is this task costing you?
Answer four questions and the estimate updates as you go. Every assumption behind the number is listed below, because an estimate you cannot check is not worth much.
ROI calculator
What would automating this task save you?
Enter what the task costs you today. The estimate updates as you change the inputs, and every assumption behind it is listed below.
Prefer to talk through your own numbers? Get in touch
hours per week
Estimate based on conservative assumptions: 48 working weeks, your selected automation share, fully loaded labor cost. Figures in USD.
How this is calculated
- Hours recovered = hours per week x people x 48 working weeks x the automation share you selected.
- 48 working weeks rather than 52, to leave room for holidays and leave.
- An annual salary is converted to an hourly rate over 2,080 hours (40 hours x 52 weeks).
- The hourly rate is multiplied by 1.3 for fully loaded labor cost. That is the midpoint of the commonly cited 1.25 to 1.4 range.
- Hours are rounded to the nearest 10 and savings to the nearest $500, because the inputs are estimates and false precision helps nobody.
- All figures are in USD. The estimate covers recovered time only, and not the cost of building or running the automation.
How this estimate is built
The calculation
Hours recovered per year = hours per week x people x 48 working weeks x the automation share you selected. Annual savings = those hours x a fully loaded hourly cost.
If you enter an annual salary, it becomes an hourly rate over 2,080 hours, which is 40 hours across 52 weeks. That rate is then multiplied by 1.3 to approximate the fully loaded cost of employing someone: payroll taxes, benefits, equipment and the rest.
Why the assumptions lean conservative
The defaults are chosen to under-promise. 48 working weeks instead of 52 leaves room for holidays and leave. The 1.3 multiplier is the midpoint of the 1.25 to 1.4 range that is commonly cited, not the top of it. The automation share defaults to 50%, not 100%, because almost no process is fully removed on the first pass.
Results are rounded to the nearest 10 hours and the nearest $500. The inputs are estimates, so precision past that would be decoration. Below $2,000 a year the money figure is dropped entirely and only the hours are shown, because a saving that small should not carry a decision on its own.
What the estimate does not include
It measures recovered time only. It does not subtract what the automation costs to build, to host or to maintain, and it does not price the time your team spends changing how they work.
It also assumes the recovered hours go somewhere useful. Time freed from a task is only a saving if it is spent on something that matters or if it removes the need to hire.
How to check it against reality
Time the task for one week before trusting any figure here, including this one. Most people are surprised in one direction or the other, and a single week of real data beats a good guess.
Then rerun the estimate with what you measured. If the number still looks worth acting on, it probably is.
Questions about the estimate
How accurate is this estimate?
It is an estimate, not a quote. It is only as good as the hours you enter, which is why the assumptions behind it are listed above rather than hidden. Treat it as a way to decide whether a process is worth a closer look, not as a number to put in a budget.
Why 48 working weeks instead of 52?
Nobody works 52 weeks. 48 leaves roughly a month for holidays, leave and the weeks where the task simply does not happen. Using 52 would inflate every result by about 8%, and an inflated estimate is the fastest way to lose the argument later.
What does the 1.3 multiplier cover?
Employing someone costs more than their salary: payroll taxes, benefits, software, equipment and space. 1.3 is the midpoint of the 1.25 to 1.4 range commonly used for fully loaded labor cost. If your own finance team uses a different figure, the honest move is to use theirs.
Can automation really remove that share of the work?
Sometimes, and often not on the first attempt. The 50% default assumes half the task is repetitive enough to hand over and the rest still needs a person. Move it to 70% only if the process is genuinely rule-based end to end, and to 30% if judgment is involved at most steps.
Why did the dollar figure disappear?
Below $2,000 a year the estimate shows hours only. At that size the rounding is a large share of the number, and a money figure would imply a confidence the math does not support. The hours are still worth knowing.
What happens to the numbers I enter?
The calculation runs in your browser. If you have accepted analytics, we receive the hours, headcount and automation share you selected plus a range for the estimated saving, never the salary or rate itself. The full figures are sent only if you request the emailed breakdown. The privacy policy lists every field and where it goes.
Put it against a real process
If the estimate looks worth acting on, the next step is a short conversation about the specific process behind it, what it would take to automate, and whether it is the right one to start with.