tester-experiments

What problem does it solve?

Establishes statistically rigorous experiments and feature flag tests in Braze to determine whether messaging, content variants, or feature changes produce measurable improvements in user engagement and conversions.

Core Features & Use Cases

• Bucket-based A/B testing: Use Braze random_bucket_number ranges to create deterministic control and treatment cohorts for campaign and Canvas experiments.
• Feature flag experiments: Gate functionality with feature flags, define variants, and measure downstream conversion events to guide rollout decisions.
• Statistical integrity: Follow sample size/power calculations, pre-register primary metrics and fixed durations, monitor guardrail metrics, and apply multiple-comparison corrections when needed.
• Use case: Validate a new messaging template or UI feature by running a controlled Canvas experiment with clearly defined primary metric and a documented ship/iterate decision process.

Quick Start

Run an A/B test in Braze using random bucket slices (for example control 0–4999 vs treatment 5000–9999) with a declared primary metric, a fixed analysis window, and a sample-size plan.