How to Calculate Conversion Lift
Translate treatment and reference conversion rates into absolute and relative lift without overstating the result.
The Short Version
Translate treatment and reference conversion rates into absolute and relative lift without overstating the result. A practical way to begin is to calculate the conversion rate separately for the reference and treatment groups, then subtract the reference rate from the treatment rate to obtain absolute lift in percentage points. Finish by making sure you check whether the plausible range includes effects too small to matter operationally. The full walkthrough below explains each part in order.
Is This Guide for You?
Researchers and managers interpreting binary outcomes such as clicked, signed up, purchased, donated, or renewed.
Follow These Steps
Work through these 6 steps at your own pace. The wording is intentionally practical, and you can return to the checklist at the end when you are ready to review your work.
- Calculate the conversion rate separately for the reference and treatment groups.
- Subtract the reference rate from the treatment rate to obtain absolute lift in percentage points.
- Divide the absolute difference by the reference rate to obtain relative lift when the reference rate is not zero.
- Report the sample size and 95% confidence interval for the difference.
- Translate the estimate into expected additional conversions at a realistic traffic level.
- Check whether the plausible range includes effects too small to matter operationally.
What This Helps You Accomplish
Absolute and relative lift answer different questions. A change from 10% to 12% is a 2 percentage-point increase and a 20% relative lift; reporting only the larger-looking number can mislead decision-makers.
A Quick Confidence Check
- Label percentage points and percentages correctly.
- Show both group rates.
- Include uncertainty.
- Connect lift to realistic volume.
- Do not treat statistical significance as business importance.
Common Questions
Can I use this guide if I am new to this?
Yes. Researchers and managers interpreting binary outcomes such as clicked, signed up, purchased, donated, or renewed. Follow the steps in order, start with a small test, and use the final checklist before you field or report anything important.
What is the simplest way to get started?
Begin by calculate the conversion rate separately for the reference and treatment groups. Next, subtract the reference rate from the treatment rate to obtain absolute lift in percentage points. You do not need to perfect every setting before running a small preview or pilot.
How do I know when I am ready?
Use the confidence check above. In particular: Label percentage points and percentages correctly. Show both group rates. Include uncertainty. Connect lift to realistic volume. When the decision is consequential, keep your study documentation and ask a qualified colleague or methods reviewer to examine the design as well.
Related Guides
Research artifact: sample data, codebook, and scriptsResearch Pathways
Product names and fielding platforms mentioned in this guide belong to their respective owners. Conjoint Survey is an independent academic research tool and is not affiliated with, sponsored by, or endorsed by those services.