What sample size is needed for MVT?

MVT sample size requirements depend on the number of combinations being tested and the effect size you need to detect.

A basic rule: each combination needs enough traffic to achieve significance independently. If an A/B test needs 5,000 recipients per variant, an 8-combination MVT needs approximately 40,000 total.

Factors affecting requirements:

Number of combinations: 2 variables × 2 options = 4 combinations. 3 variables × 2 options = 8 combinations. It multiplies quickly.

Baseline conversion rate: Lower baseline rates need larger samples.

Minimum detectable effect: Detecting small differences requires larger samples than detecting large ones.

Desired confidence level: Higher confidence (99% vs 95%) increases requirements.

Use MVT calculators specific to factorial designs rather than simple A/B calculators.

Insufficient sample size is the most common MVT mistake. If you cannot meet the requirements, run sequential A/B tests instead.