UX ROI calculator: when does a UX investment pay off?

UX ROI calculator

How it's calculated

The tool works with five inputs and calculates in four steps. All formulas are here. If you like, build the same calculation yourself in a spreadsheet.

1. New conversion rate. In percentage points: new rate = current rate + lift. Relative in %: new rate = current rate × (1 + lift / 100).
2. Extra revenue a month = visits a month × (new rate − current rate) × average order value.
3. Payback period (months) = investment / extra revenue a month.
4. ROI a year (%) = (12 × extra revenue a month − investment) / investment × 100.

The usual definition of ROI is: (gain from investment − cost of investment) / cost of investment. The payback period comes from dividing the investment by the annual net cash flow.

The tool applies the same logic per month. I took both definitions from the Wikipedia articles.

Both calculations have known limits. ROI is not time-adjusted, so it counts a dollar today and a dollar in a year the same. The payback period ignores the revenue after the investment has been paid back.

One limit comes from me: the tool takes the extra revenue as the gain, not the profit. If you want to calculate with profit, enter your gross margin per order instead of the average order value. The result is more cautious and, in my opinion, more honest.

Three examples, calculated by hand

I checked the tool with these three scenarios. I made the numbers up; none comes from a real company. They're only there to test the formula.

Example 1. 20,000 × 0.005 = 100 extra orders. 100 × $75 = $7,500. 15,000 / 7,500 = 2 months. Extra revenue a year $90,000, ROI (90,000 − 15,000) / 15,000 = 500%.

Example 2. A relative 10% on a rate of 1.5% makes the rate 1.65%. 50,000 × 0.0015 = 75 extra orders, 75 × $40 = $3,000. Payback 8,000 / 3,000 ≈ 2.7 months. ROI (36,000 − 8,000) / 8,000 = 350%.

Example 3. 5,000 × 0.002 = 10 extra orders, 10 × $30 = $300. Payback 6,000 / 300 = 20 months. Extra revenue a year $3,600, ROI (3,600 − 6,000) / 6,000 = −40%.

The third example goes negative on purpose. With little traffic, even a good improvement may not pay off in the first year. Seeing that before the investment is better than after.

The claim that “$1 in UX returns $100”

I often hear this sentence in UX presentations. I couldn't find and read the primary source, so I treat it as an unsupported claim.

One of the texts that repeats the claim links to the Forrester report “The ROI Of Design Thinking: Part 1, Overview” as its source. According to Forrester's page, the report is dated May 16, 2019. The full text is paywalled, and the public summary has no $100 figure.

I don't defend a budget with a multiplier whose source can't be read. An average from other companies doesn't know your traffic either. Calculate with your own numbers on this page.

Where to get the expected lift

The weakest input in the calculation is the expected lift. That's why the tool doesn't suggest a value. You have to supply the number yourself.

If you have an earlier test, use it. The A/B test calculator shows whether a difference is significant and gives the confidence interval. You can enter the lower end of the interval here as a pessimistic scenario.

If there is no test, calculate at least twice: once with a pessimistic and once with an optimistic lift. If the payback is still acceptable in the pessimistic scenario, the decision is easy.

Check whether you measure your current rate correctly with the conversion rate guide.

My note

When it comes to ROI, I don't defend an optimistic picture. I build the pessimistic scenario first and want the decision to rest on it.

The estimate of the lift is only as solid as your knowledge of where you lose people. In ecommerce UX work I look at the cart and the payment step first. A lift entered without knowing where the loss is, is only a hope.

If you want to find out which flow you lose people in, I can review up to 10 flows in UX Audit Pro. I rank the fixes by impact and effort. That order shows which one to enter in this calculation first.