What would automating this actually return?
Five numbers from you, arithmetic from us. No industry benchmark, and no success rate we claim to have measured, and the time a human spends approving gated steps is deducted rather than ignored — which is the line item most vendor calculators leave out.
How many times this workflow runs in a typical week.
Include the switching cost: opening tabs, finding records, waiting on pages.
Salary plus overhead for whoever does this today — blended, if it's several people.
Supervised automation doesn't go to zero. This is the time an approver spends on a gated step, and it's deducted from the saving.
Estimated 5.1 hrs reclaimed per week, $12,012 of labor cost per year.
- Time on this workflow today
- 8.0 hrs/wk
- Cases still handled fully by a person
- 12/wk
- Reviewing and approving gated steps
- 0.5 hrs/wk
- Time on this workflow after
- 2.9 hrs/wk
Read the assumptions before you quote the number.
- Compute only from the fields as they stand. The starting values are a worked example to overwrite — 70% is not a rate we measured.
- Deduct approval and review time on automated cases from the saving.
- Keep the cases that don't automate at their full manual cost.
- Report no saving when reviewing a case would cost as much as doing it.
- Warn you when the weekly saving is too small to justify the upkeep.
- Include what any software costs — including ours, which is on design-partner terms rather than a public price.
- Include the one-off effort of writing the brief, testing it, and tightening it in week one.
- Claim an automation success rate. That slider is your assumption, and we don't publish a measured figure.
- Model error costs, either direction — a mistake caught at a gate, or one that gets through.
- Promise the reclaimed hours turn into budget. They turn into capacity, which is worth less on a spreadsheet and often more in practice.
Size looking worth it? The next question is whether the workflow is the right shape— frequency and duration say it's big enough, but not that an agent can do it. The readiness check covers that, and the manual-versus-automated guide walks through the reasoning behind these numbers.
Questions about this calculator
How do you calculate automation ROI?
Multiply cases per week by minutes per case to get the time the workflow costs today. Split the cases by the share that automates cleanly: the rest still cost their full time, and the automated ones cost only the minutes a reviewer spends approving them. The difference is the weekly saving; multiply by 52 for the year and by the fully loaded hourly rate for the labor cost. That's the whole formula. One guardrail on top: we never display a negative saving, so if reviewing costs more than doing, the tool reports no saving and says the gate is likely misplaced.
Why does this calculator ask for review time?
Because supervised automation doesn't go to zero, and a calculator that ignores that is flattering rather than useful. If an approver spends a minute per case confirming a gated step, that minute is a real cost and it's deducted here. When review time approaches the time the task took manually, the tool reports no saving — which is a signal the gate is in the wrong place.
What automation success rate should I assume?
One you can defend, and lower than you'd like. We deliberately don't supply a number: we have no published measurement of ours, so quoting one would be marketing rather than a forecast. The starting value here is 70%, marked as an assumption for you to change. If a vendor gives you a savings percentage before asking about your workflow, treat it as a sales figure.
Does this include the cost of the software?
No, and it says so plainly. This estimates the value side only — hours and labor cost returned. Boring AI is in early access with design-partner terms rather than a published price list, so putting our cost into the formula would mean inventing one. Compare the output against whatever a vendor actually quotes you.
What counts as a fully loaded hourly cost?
Salary plus employer taxes, benefits, and overhead for whoever does the work today — typically meaningfully more than base salary divided by working hours. If the work is spread across several people at different rates, use a blended figure weighted toward whoever does most of it.
Is the saved time real, or does it just move?
It moves, and being honest about that matters. Reclaimed hours become capacity, not a line item you can bank — the value shows up as work that was being deferred getting done, or a team absorbing growth without adding headcount. Treat the dollar figure as the size of the opportunity, not as money that appears in a budget.
Worth it on paper? Let's do it for real.
Request access and tell us the workflow you just costed — or ask about white glove and our team will build and run it with you.