Plug in your traffic, baseline conversion rate and the lift you expect to win. This calculator projects extra conversions, added revenue and the return on every dollar spent running the experiment.
| Metric | Value |
|---|---|
| Variant conversion rate | — |
| Added conversions / month | — |
| Added revenue / month | — |
| Gross revenue over horizon | — |
| Payback period | — |
Most ROI sheets stop at "extra revenue." This tool models the full economics of an experiment so you can decide whether the test is worth running before you build it. The engine runs five chained steps:
First it converts your baseline rate into the variant rate with variant = baseline × (1 + lift/100). The difference in rates, multiplied by monthly visitors, gives added conversions per month: extra = visitors × (variant − baseline). Note this uses the relative lift convention — a 12% lift on a 3.2% baseline yields a 3.584% variant rate, not 15.2%, which is the single most common ROI miscalculation teams make.
Added monthly revenue is extra × value per conversion. We then project it across your benefit horizon (months the winner stays live) to get gross revenue, subtract the one-time test cost to get net revenue, and compute ROI = net ÷ cost × 100. Finally the payback period — cost ÷ monthly revenue — tells you how many months of the rolled-out winner it takes to repay the experiment. A payback under one month with a positive ROI is the green-light signal most growth teams use.
The information this adds over a plain "uplift × AOV" estimate is the opportunity-cost frame: by holding test cost and horizon explicit, you can compare a cheap copy test against an expensive checkout rebuild on the same ROI axis. Pair it with a power analysis — a real lift only counts once it clears statistical significance, so size your sample before you trust the projection. ROI assumes the measured lift holds after rollout; in practice apply a discount for novelty decay and regression to the mean.