A/B Testing Screen Resolution (Annual) Calculator
Estimate the annual test traffic available for a screen resolution and whether it is enough to detect a meaningful conversion-rate change.
Overview
Use this A/B Testing Screen Resolution (Annual) Calculator to estimate how much annual traffic is available for a specific screen resolution and whether that traffic can support a conversion experiment. Enter your annual traffic, the resolution segment's share, baseline conversion rate, desired detectable change, and test settings.
How it works
The calculator first estimates annual visitors for the selected screen resolution by multiplying total annual traffic by the resolution's traffic share. It then applies the portion of traffic available for experimentation. The required sample uses a standard approximation for comparing conversion rates: a smaller baseline conversion rate or smaller detectable effect generally requires more visitors. Required visitors per variant are multiplied by the number of variants and compared with annual testable traffic.
How to use this calculator
- 1Enter your estimated annual visitor or session volume.
- 2Add the percentage of traffic using the screen resolution you want to assess.
- 3Choose the share of that segment you can send into the test.
- 4Enter the segment's current conversion rate and the minimum relative change you want to detect.
- 5Set the number of variants, confidence z-score, and power z-score.
- 6Review the visitor requirement and estimated number of tests supported annually.
Example Calculation
Annual visitors
1000000
Screen resolution traffic share
25%
Traffic allocated to testing
80%
Baseline conversion rate
4%
Minimum detectable effect
20%
Number of variants
2
Confidence z-score
1.96
Statistical power z-score
0.84
Annual testable visitors
200,000 visitors
With 1,000,000 annual visitors, a 25% resolution share, and 80% test allocation, about 200,000 visitors are available annually. Detecting a 20% relative change from a 4% baseline needs about 18,816 visitors in total, so the segment could support about 10 similar non-overlapping tests per year.
Frequently asked questions
What is a screen-resolution A/B test?
It is an experiment analysed for visitors using a particular screen resolution or resolution group, such as a desktop or mobile viewport segment.
Why calculate traffic by screen resolution?
A page change can perform differently across device layouts. Segment-level traffic helps determine whether there is enough data to test or analyse that experience reliably.
What does minimum detectable effect mean?
It is the smallest relative conversion-rate change you want the test to have a reasonable chance of detecting. Smaller effects require larger samples.
Which confidence and power values should I use?
A common starting point is a confidence z-score of 1.96, representing about 95% two-sided confidence, and a power z-score of 0.84, representing about 80% power.
Why might my actual test need more traffic?
Uneven traffic splits, multiple variants, seasonal changes, visitor overlap, low-quality data, and multiple metrics can all make the practical requirement higher.
Can I combine similar screen resolutions?
Combining resolutions can increase sample size, but only do so when the layouts and user experiences are similar enough for a combined analysis to be meaningful.
Explore Related Calculators
Assumptions and warnings
Assumptions
- Traffic share for the selected screen resolution remains broadly stable throughout the year.
- Visitors are assigned evenly between variants.
- The calculation uses an approximate two-sided conversion-rate sample-size method.
- Tests are treated as non-overlapping and use independent visitors.
- Results are planning estimates; actual traffic, conversion rates, and experiment quality can vary.
Warnings
- This calculator provides an experimental-planning estimate only, not business, financial, or professional advice.
- Low-volume resolution segments, repeat visitors, unequal allocation, multiple comparisons, and data-quality issues can increase the sample size needed.