A/B Testing Website Load Time Calculator Examples
Worked examples showing how monthly traffic and page-load differences affect aggregate visitor waiting time.
These examples show how a per-page speed change can scale across different traffic volumes, page-view patterns, and test outcomes. The figures estimate aggregate waiting time and are not measures of business impact.
High-traffic website with a major speed improvement
A website receives 100,000 visitors per month, and visitors view 3 pages on average.
Input Summary
Monthly visitors
100,000
Page views per visitor
3
Variant A load time
3.2 seconds
Variant B load time
1.2 seconds
Calculation Breakdown
- 1Estimate page views100,000 × 3300,000 page views
- 2Find per-page difference3.2 − 1.22.0 seconds
- 3Calculate total seconds300,000 × 2.0600,000 seconds
- 4Convert to hours600,000 ÷ 3,600166.7 hours
Result Summary
Calculate total seconds
600,000 seconds
A/B Testing Website Load Time Calculator
Variant B saves an estimated 166.7 aggregate visitor waiting hours per month and is 62.5% faster than variant A.
Small content site with a modest speed improvement
A content site has 8,000 monthly visitors, each viewing 2.5 pages on average.
Input Summary
Monthly visitors
8,000
Page views per visitor
2.5
Variant A load time
2.5 seconds
Variant B load time
2.0 seconds
Calculation Breakdown
- 1Estimate page views8,000 × 2.520,000 page views
- 2Find per-page difference2.5 − 2.00.5 seconds
- 3Calculate total seconds20,000 × 0.510,000 seconds
- 4Convert to hours10,000 ÷ 3,6002.8 hours
Result Summary
Calculate total seconds
10,000 seconds
A/B Testing Website Load Time Calculator
Variant B saves an estimated 2.8 aggregate visitor waiting hours per month and reduces average load time by 20.0%.
E-commerce site with deep browsing sessions
An e-commerce site receives 50,000 visitors monthly, with 6 page views per visitor.
Input Summary
Monthly visitors
50,000
Page views per visitor
6
Variant A load time
2.8 seconds
Variant B load time
2.0 seconds
Calculation Breakdown
- 1Estimate page views50,000 × 6300,000 page views
- 2Find per-page difference2.8 − 2.00.8 seconds
- 3Calculate total seconds300,000 × 0.8240,000 seconds
- 4Convert to hours240,000 ÷ 3,60066.7 hours
Result Summary
Calculate total seconds
240,000 seconds
A/B Testing Website Load Time Calculator
Variant B saves about 66.7 aggregate visitor waiting hours per month and is 28.6% faster than variant A.
Variant B is slower than the control
A website has 25,000 monthly visitors and 4 page views per visitor.
Input Summary
Monthly visitors
25,000
Page views per visitor
4
Variant A load time
1.8 seconds
Variant B load time
2.3 seconds
Calculation Breakdown
- 1Estimate page views25,000 × 4100,000 page views
- 2Find per-page difference1.8 − 2.3−0.5 seconds
- 3Calculate total seconds100,000 × −0.5−50,000 seconds
- 4Convert to hours−50,000 ÷ 3,600−13.9 hours
Result Summary
Calculate total seconds
−50,000 seconds
A/B Testing Website Load Time Calculator
Variant B adds an estimated 13.9 aggregate visitor waiting hours per month and is 27.8% slower than variant A.
How to Read Your Results
A positive load-time difference means variant B is faster; a negative difference means it is slower.
Monthly visitor time saved is an aggregate total across page loads, not the time one individual visitor necessarily experiences.
The improvement percentage is measured relative to variant A's average load time.
Use the same load-time metric and comparable traffic conditions for both variants.
Scale the result down if only a portion of traffic or only certain pages will receive variant B.
Assumptions & Important Notes
- Each example assumes all calculated page views would use variant B after rollout.
- Load-time values are comparable averages measured under consistent conditions.
- Monthly traffic and page-view behavior remain stable during the estimate period.
- The calculation measures wait-time differences only, not behavior or commercial outcomes.
Related Examples
Frequently Asked Questions
Can I use the calculator for a partial rollout?
Yes. Estimate the page views that will actually receive variant B rather than using all monthly traffic, or calculate the full result and multiply it by the expected rollout share.
Why can two sites with the same visitor count get different results?
They may have different page views per visitor, different per-page speed changes, or both. The monthly page-load total drives the aggregate result.
Is a 0.5-second improvement meaningful?
The calculator shows its aggregate time effect. Whether it is meaningful for a particular site depends on the audience, page types, measurement quality, and other test results.
What if visitors view different types of pages?
Use a weighted average load time and weighted average page-view mix where possible, or calculate important page groups separately.
Can the calculator compare mobile and desktop performance?
Yes, but calculate each audience separately when their load times or page-view patterns differ, then combine the results if needed.
Ready to calculate your own result?
Use the live calculator with your own inputs, timing, and preferences.