A/B Testing Website Load Time Calculator Examples
Explore worked examples of website speed A/B tests with different traffic levels, conversion changes, and conversion values.
These examples show how a load-time test can be translated into speed savings, relative conversion uplift, and a planning estimate for a full rollout. They illustrate calculations, not proof that page speed caused the outcome.
How to Read Your Results
Load time saved is positive when the variant is faster and negative when it is slower.
Speed improvement compares the time saved with the control load time.
Conversion-rate change is stated in percentage points; relative uplift compares that change with the control rate.
Projected conversions and revenue apply the observed rate difference to all monthly visitors, not just test traffic.
A positive projection is an estimate that should be considered alongside test quality and uncertainty.
Assumptions & Important Notes
- Each example assumes the measured control and variant rates are comparable.
- The full-rollout estimate assumes the observed conversion-rate difference holds across all monthly visitors.
- Conversion value is assumed to be constant for the projected period.
Related Examples
Frequently Asked Questions
What is a simple page-speed A/B test example?
If a page improves from 3.5 seconds to 2.5 seconds and conversion rises from 3.0% to 3.3%, the change is 0.3 percentage points or 10% relative uplift.
How does visitor volume affect the result?
For the same conversion-rate difference, more monthly visitors produce a larger projected number of additional conversions.
How does average conversion value affect projected revenue?
It does not change the conversion estimate, but a higher value per conversion increases the projected revenue impact.
Can a faster variant produce lower revenue?
Yes. If its observed conversion rate is lower than the control rate, the projection will show a negative conversion and revenue impact.
Ready to calculate your own result?
Use the live calculator with your own inputs, timing, and preferences.