Utility Lab · live calculation
AI ROI Calculator
Honest numbers, no vendor sales sheet. Enter people, hours saved, salary cost, tool cost and one-off setup. The calculator returns year-one savings, payback in months, three-year net value and a tier judgement on whether the case is worth taking to your board.
- Annual benefit
- £32,256
- Year-1 cost
- £17,000
- Payback
- 6 mo
- 3-year ROI
- 234%
- 3-year net value
- £67,768
These numbers stand up to a board challenge. Your remaining work is verifying the assumptions - particularly the hours-saved figure - with a two-week manual proof of concept before signing the contract.
3-year cost vs benefit
Email me my ROI case + book a stress-test
Simon receives the live numbers above plus your input assumptions, so the follow-up call goes straight to the weak points in the case rather than starting from scratch.
What gets sent with your enquiry
Tier: Strong case Annual benefit: £32,256 Year-1 cost: £17,000 Payback: 6 months 3-year ROI: 234% 3-year net value: £67,768 Inputs: - People: 10 - Hours saved per person per week: 3 - Hourly cost: £32 - Adoption: 70% - Tool annual cost: £6,000 - Implementation: £8,000 - Training: £3,000
How this is calculated
Assumptions in plain English
- • Annual benefit = people × adoption % × hours saved per week × 48 working weeks × fully-loaded hourly cost.
- • Year-1 cost = tool subscription + one-off implementation + training/change cost. Year 2 & 3 assume tool subscription only.
- • Payback = year-1 cost ÷ annual benefit, expressed in months. 3-year ROI = 3-year net value ÷ total 3-year cost.
- • Tier bands: Strong (ROI ≥ 200% AND payback ≤ 12 months), Workable (ROI ≥ 50% AND payback ≤ 24 months), Marginal (ROI ≥ 0%), otherwise Don't proceed.
- • Designed to be conservative. The number that breaks most board cases is over-stated hours saved per person per week - sanity-check it with a two-week manual pilot before signing anything.
- • All maths runs in your browser. Nothing is sent or stored unless you email yourself the result.
AI ROI calculator: common questions
How accurate is a free online AI ROI calculator compared to a bespoke business case?
It is only as accurate as the assumptions you put in, particularly the hours-saved-per-person figure - that is the number that breaks most board cases when it turns out to be optimistic. Treat the output as a directional first pass, not a final business case. Validate the hours-saved assumption with a two-week manual pilot before taking the numbers to your board or signing a vendor contract.
What adoption rate should I use in the AI ROI calculator?
Most AI tool roll-outs land a realistic adoption rate of 50-75% after six months, even when every licensed user is trained. Assuming 100% adoption is the most common way UK SMEs overstate an AI business case. If you have no prior data, start at 60% and treat anything above that as an upside case you would need evidence to defend.
Why does the calculator separate year-one cost from ongoing annual cost?
Because implementation and training are one-off costs that inflate year one and then disappear, while the tool subscription repeats every year. Blending them into a single average understates the real payback period in year one and overstates the ongoing cost from year two onwards. Splitting them out is what makes the payback-in-months and three-year ROI figures usable in a board pack.
What counts as a strong AI ROI case versus one that needs more work?
On this calculator, a strong case needs a three-year ROI of 200% or more with payback inside 12 months. A workable case needs 50% or more ROI with payback inside 24 months. Anything at or above 0% but below that is marginal - it pays for itself on paper but has no margin for unexpected friction. Below 0%, the investment loses money over three years on the assumptions entered and the use case needs re-scoping before it goes further.
A weak business case kills more AI projects than weak technology.
Bring your ROI numbers to a free 30-minute call. We'll stress-test the assumptions and decide together whether the project is worth the risk.