A/B Testing Download Time Calculator Examples
Worked examples show how page size, connection speed, and transfer efficiency affect estimated A/B test download times.
These scenarios show how the calculator estimates file-transfer time when two versions have different download sizes. They are illustrative estimates, not measurements of complete user-perceived page speed.
Landing page image optimization
Compare a 4 MB original page with a 2.5 MB optimized version on a mid-range connection.
Input Summary
Version A size
4 MB
Version B size
2.5 MB
Connection speed
20 Mbps
Transfer efficiency
75%
Calculation Breakdown
- 1Effective speed20 × 0.7515 Mbps
- 2Version A time(4 × 8) / 152.13 seconds
- 3Version B time(2.5 × 8) / 151.33 seconds
- 4Time saved2.13 - 1.330.80 seconds
Result Summary
Time saved
0.80 seconds
A/B Testing Download Time Calculator
Version B is estimated to download 0.80 seconds sooner than Version A.
Mobile-focused product page
Compare a 2 MB page with a 1.2 MB version for visitors with an effective low-speed connection.
Input Summary
Version A size
2 MB
Version B size
1.2 MB
Connection speed
8 Mbps
Transfer efficiency
70%
Calculation Breakdown
- 1Effective speed8 × 0.705.6 Mbps
- 2Version A time(2 × 8) / 5.62.86 seconds
- 3Version B time(1.2 × 8) / 5.61.71 seconds
- 4Relative difference(1.15 / 2.86) × 10040.0%
Result Summary
Relative difference
40.0%
A/B Testing Download Time Calculator
Version B saves about 1.15 seconds and is estimated to download 40.0% faster than Version A.
Small script bundle change
Compare a 1.1 MB version with a 0.9 MB version on a fast connection.
Input Summary
Version A size
1.1 MB
Version B size
0.9 MB
Connection speed
50 Mbps
Transfer efficiency
85%
Calculation Breakdown
- 1Effective speed50 × 0.8542.5 Mbps
- 2Version A time(1.1 × 8) / 42.50.21 seconds
- 3Version B time(0.9 × 8) / 42.50.17 seconds
- 4Time differenceabs(0.21 - 0.17)0.04 seconds
Result Summary
Time difference
0.04 seconds
A/B Testing Download Time Calculator
Version B is estimated to save roughly 0.04 seconds of transfer time.
Heavy campaign page versus streamlined variant
Compare a 7 MB experience with a 3.5 MB streamlined alternative on a typical broadband connection.
Input Summary
Version A size
7 MB
Version B size
3.5 MB
Connection speed
25 Mbps
Transfer efficiency
80%
Calculation Breakdown
- 1Effective speed25 × 0.8020 Mbps
- 2Version A time(7 × 8) / 202.80 seconds
- 3Version B time(3.5 × 8) / 201.40 seconds
- 4Time difference2.80 - 1.401.40 seconds
Result Summary
Time difference
1.40 seconds
A/B Testing Download Time Calculator
Version B is estimated to download 1.40 seconds sooner, a 50.0% reduction relative to Version A.
How to Read Your Results
Compare the two estimated times under identical network assumptions; the lower value indicates the lighter transfer burden.
The time difference is expressed in seconds and reflects file transfer only.
The percentage difference is measured relative to Version A, so Version A is the baseline.
Use audience-representative connection speed and efficiency inputs when evaluating a test.
Consider the result alongside real-user metrics and controlled performance measurements.
Assumptions & Important Notes
- Both variants are tested with the same connection speed and transfer-efficiency setting.
- All entered MB are assumed to be downloaded during the visit.
- The examples use decimal MB and Mbps for a consistent comparison.
- Caching, server response, and browser processing are outside these transfer-time examples.
Related Examples
Frequently Asked Questions
Can I use these examples for image optimization tests?
Yes. Enter the total downloaded size of each image or page version and use the connection assumptions relevant to the audience.
Why do savings look larger in the mobile example?
A fixed size reduction takes longer to transfer on a slower effective connection, increasing the estimated time saved.
Does halving page size always halve download time?
In this formula, yes, when effective connection speed and all other assumptions remain unchanged.
Are these examples full page-load measurements?
No. They show estimated transfer time and exclude processing, rendering, and other stages of loading.
Ready to calculate your own result?
Use the live calculator with your own inputs, timing, and preferences.