AZ400-Q: separate A/B evaluation from deployment
Product wants to compare two checkout experiences for randomly selected users while both versions are deployed. Which approach is most appropriate?
- A) Use a feature-flag/experimentation mechanism that assigns cohorts and records comparable usage/outcome telemetry. ✓
- B) Deploy one version to production and the other to a developer laptop.
- C) Create two unrelated Git branches and compare commit counts.
- D) Add a resource lock.
Answer: A. Cohort assignment and telemetry allow an evidence-based comparison without treating deployment itself as the experiment.
Why not the others: Separate environments, branches, and locks do not create controlled user experimentation.