Request a demo Talk to an architect

The formula in one line

Annual manual hours × fully loaded hourly cost gives the current labour cost. Subtract the human hours that remain after automation (cases not attempted, reviews, failures, rework) to get hours returned; their value × your cashable share is the realized benefit. Realized benefit − annual operating cost = net annual benefit, and implementation cost ÷ monthly net benefit gives payback. No email is needed: results update as you type.

How this automation ROI calculator works

It starts from the work you do today: how many cases, how long each takes, or how many people spend how much of their time on the process. It then follows every case through one outcome: not attempted, completed straight through, completed after human review, or failed and finished by a person. The hours that remain are compared with today, valued at your fully loaded hourly cost, and set against the operating cost of the automation you choose.

RPA vs AI agents: what changes in the cost

RPA cost is mostly platform, robot runtime, hosting and maintenance. AI-assisted and agent-led automation add a usage cost per case: model tokens, tool or API calls and retries, which you can enter as one figure per case or stage by stage. The calculator never gives agentic approaches a higher automation rate: coverage, success and review are always your own assumptions.

Capacity value is not cash

Hours returned are valued twice, separately: the gross capacity value of all freed time, and the realized financial benefit, which is only the cashable share you enter (fewer hires, less overtime or contractor spend). Net benefit, payback and ROI use the realized benefit only.

How do I calculate automation ROI?

Work out today’s annual manual hours and their fully loaded cost, estimate the human hours that remain after automation (cases not attempted, reviews, failures and rework), and value the difference. Multiply by the share that is cashable, subtract the annual operating cost of the automation, and compare the result with the implementation cost to get payback, first-year ROI and the 3-year net benefit.

How is an RPA ROI calculation different from an AI agent ROI calculation?

The labour side is identical. The cost side differs: RPA is mostly fixed platform and runtime cost, while AI agents add a variable cost per case from model tokens, tool calls and retries. AI agents may also need more human review. Enter both honestly and compare the approaches with the same financial definitions.

What does automation coverage mean?

Coverage is the share of all cases that automation attempts. Of the attempted cases, some finish straight through, some finish after a human review and some fail and are handled manually. These three shares always add up to 100% of attempted cases, so no manual effort is counted twice.

Why is my net benefit "Not calculated"?

The calculator never fills a missing input with zero or a guess. If a figure depends on an input you have not entered, such as the cashable share, the loaded cost per employee or an exchange rate for AI prices in another currency, it shows "Not calculated" and names the field it needs.

Are AI model prices built in?

No. Model prices change often and differ by vendor and contract, so you enter your own price per million input and output tokens, or your own cost per case. If those prices are in US dollars and you report in another currency, you enter the exchange rate yourself.

Are these results a guarantee of savings?

No. The results are estimates based only on the assumptions you enter. Validate them with a pilot on real cases before committing to a business case.

Assess automation fit for your process RPA platform Agentic automation Building an automation business case your CFO will trust