Power & Significance Tooling

note

Power and significance tooling is an enterprise feature available as part of Marqo Enterprise. Contact your Marqo account manager to enable this feature.

Power analysis tells you how many sessions a test needs before it can detect the effect you care about; significance testing tells you whether an observed difference between variants is real or noise.

Marqo does not currently provide sample-size calculators, lift estimators, or guardrail-metric automation. When you run an A/B test, decide the minimum detectable effect and required sample size up front with your own tooling, and read per-variant results from analytics using the pixel_experiment labels described in the Analytics API.

Best Practices

1. Calculate upfront: Determine sample sizes before starting tests
2. Set realistic MDE: Choose meaningful minimum detectable effects
3. Run long enough: Cover full weekly cycles before drawing conclusions
4. Account for multiple comparisons: Adjust for multiple variants

Related Topics

• A/B & Multivariate Testing - Run tests
• Analytics & Reporting - Analyze results