Framework · Measurement · 3 min read
UX ROI
A conservative estimate of what better UX saves in productivity, errors, training, support and development, with every assumption stated, so leaders can weigh UX work like any other investment.
Time saved × users × frequency × loaded hourly cost.
Often the largest lever for internal toolsErrors avoided × cost per error: rework, claims, penalties.
Typically: fewer corrected bookingsShorter onboarding time × number of new hires.
Typically: days to first solo taskTickets avoided × cost per ticket.
Data often already existsLess late rework, faster delivery through the design system.
Typically: fewer change requests after releaseROI = (estimated yearly benefit − cost of the UX work) ÷ cost of the UX work
Use it when
- You need to argue for UX headcount or a redesign in a budget discussion.
- A stakeholder asks "What do we get for this?" about a UX project on an internal tool.
- You want to show the value of past UX work in a quarterly or yearly review.
How to use it
- Pick the levers that apply. Choose from productivity, error reduction, training, support and development. For internal tools, productivity and errors are usually the strongest. Output: 1–3 levers for this project.
- Get a measured baseline. Use real data where you can: time on task from a benchmark, error counts, support tickets per category, onboarding time. Output: current values with their source.
- Estimate the change conservatively. Use a measured improvement from a test, or a cautious estimate. When unsure, take the lower number. Output: an expected change per lever, with its basis.
- Turn it into money. Productivity: time saved × users × frequency × hourly cost. Errors: errors avoided × cost per error. Training: reduced onboarding time × new hires. Support: tickets avoided × cost per ticket. Use the loaded cost of employee time (salary plus benefits and overhead; ask finance). Output: a yearly benefit per lever.
- Compare with the cost and state the assumptions. ROI = (benefit − cost) ÷ cost. List every assumption next to the number, and say what is measured and what is estimated. Output: a one-page ROI estimate.
Worked example
A redesign of the booking form on an internal operations platform (fictional numbers, with assumptions stated).
| Item | Value | Basis |
|---|---|---|
| Time saved per booking | 20 seconds | Benchmark, median before vs after |
| Users × bookings per day | 120 × 25 | Usage report |
| Working days per year | 220 | Assumption |
| Loaded hourly cost | 45 | From finance |
| Raw productivity benefit | about 165,000 per year | 20 s × 25 × 120 × 220 ÷ 3600 × 45 |
| Conservative benefit (50%) | about 82,500 per year | Assumes only half the saved time becomes useful work |
| Cost of UX and dev work | 40,000 | Project estimate |
| ROI | about 106% in year one | (82,500 − 40,000) ÷ 40,000 |
The lead presents the conservative line, shows the raw line as an upper bound, and lists each assumption.
Watch out for
- The big number. Multiplying optimistic guesses gives an impressive figure that nobody trusts. Cut it down and show why.
- Hidden assumptions. If finance can't see how you got the number, they will discard it. Show the formula and the inputs.
- Salary instead of loaded cost. Salary alone undervalues time. Use the loaded cost from finance.
- Estimated presented as measured. Say clearly what you implemented, what you measured and what you only estimate.
Template
Copy this into your business case and fill only the levers that apply.
Project: Prepared by: Date:
Lever Formula Inputs (source) Yearly value
Productivity time saved x users x frequency x hourly cost
Errors errors avoided x cost per error
Training onboarding time saved x new hires x hourly cost
Support tickets avoided x cost per ticket
Development rework avoided (hours) x hourly cost
Total benefit (raw):
Conservative adjustment and why:
Total benefit (conservative):
Cost of the work:
ROI = (benefit - cost) / cost =
Assumptions:
Measured vs estimated: