AI agent ROI calculator
Estimate the hours and money a process automation would save, and how long it would take to pay for itself. Every assumption is visible and editable — including the ones that make the number smaller.
Runs entirely in your browser — nothing is sent to us
Estimated annual saving
—
Net of build and running costs, in year one.
- Payback
- —
- Hours freed / month
- —
The workings
- Current monthly cost
- —
- Monthly cost after automation
- —
- Monthly running cost
- —
- Net monthly saving
- —
- One-off build cost
- —
What this number is not
It assumes freed capacity has somewhere to go. Hours saved only become money if that time is redeployed onto work with value, or if it absorbs growth you would otherwise have hired for. If the time is simply absorbed, the saving is real but invisible on your P&L.
It excludes change management. Training, process redesign, and the weeks where people run both the old and new way in parallel are real costs this does not model.
It assumes a steady state. Month one will not hit your target automation rate. Expect a ramp as evaluation data accumulates and review thresholds loosen — which typically means the first quarter underperforms this figure and later ones exceed it.
The automatable share is the number that matters most. It is also the one people are most optimistic about. If you are unsure, halve your first instinct and see whether the case still holds — if it does, you have a robust project.
Want this checked against your actual numbers?
These figures come from general assumptions. A scoping call replaces them with yours.
Other tools
AI readiness assessment
Twelve questions about your data, processes and governance. Get a score, an honest read on where you stand, and the specific things to fix first.
LLM API cost calculator
Monthly and annual API cost for your token volumes, compared across models, including caching and batch discounts. Every price is editable.
App store readiness checker
Work through the requirements that actually cause store rejections, with the blocking ones separated from the nice-to-haves.