The Experimentation Maturity Model describes five stages that teams progress through as they build experimentation capability. Understanding where you are helps you focus on the right improvements.
Most teams start at Stage 1 or 2. Leading organizations operate at Stage 4 or 5. The goal isn't to rush to Stage 5—it's to make deliberate progress toward the next level.
Stage 1: Ad Hoc
Experiments happen occasionally, driven by motivated individuals rather than process. There's no standardized approach, and results are shared informally if at all.
Sporadic testing
Experiments happen when someone champions them, not as standard practice.
No clear process
Each experiment is set up differently. No templates or standards exist.
Results go nowhere
Findings are shared informally or not at all. Decisions rarely reference data.
How to advance
Establish basic infrastructure. Run your first intentional, documented experiment.
Stage 2: Emerging
A dedicated person or team runs experiments with some structure. Process exists but isn't always followed. Results are shared but not widely consumed.
Dedicated ownership
Someone is responsible for experimentation, usually on the data or growth team.
Inconsistent process
Templates exist but aren't always used. Quality varies between experiments.
Limited audience
Results reach a small group. Most of the organization remains unaware.
How to advance
Increase visibility. Broadcast experiments to the whole team. Make results impossible to miss.
Stage 3: Scaling
Multiple teams run experiments regularly. Clear process guides hypothesis to decision. Experiment velocity is tracked as a key metric.
Cross-team adoption
Product, growth, and engineering all run experiments. It's not just one team's job.
Standardized process
Hypothesis templates, launch checklists, and decision protocols are followed consistently.
Velocity tracking
Experiments per month is a team metric. Leadership reviews experiment activity.
How to advance
Embed experimentation in product development. Make it the default, not an add-on.
Stage 4: Optimizing
Experimentation is core to how products are built. Every major feature is tested. Meta-analysis reveals patterns across experiments.
Experimentation by default
New features launch as experiments. Shipping without testing requires justification.
Rich experiment history
All past experiments are documented and searchable. Teams learn from prior tests.
Pattern recognition
Meta-analysis identifies what types of changes typically succeed or fail.
How to advance
Expand beyond product. Bring experimentation to marketing, operations, and other functions.
Stage 5: Leading
Experimentation drives company strategy. Real-time testing informs all decisions. The organization is recognized externally for experimentation excellence.
Strategic experimentation
Company strategy is informed by experiment results. Testing shapes major decisions.
Sophisticated methods
Advanced statistical techniques (multi-armed bandits, sequential testing) are standard.
External recognition
The company publishes learnings, speaks at conferences, and attracts experimentation talent.
Continuous improvement
Even at this stage, teams seek new methods and higher velocity.
How to Apply This Framework
Assess your current stage
Be honest about where you are. Most teams overestimate by one level. Use the characteristics to determine your true stage.
Identify your stage blockers
What's preventing you from reaching the next level? Is it process, tooling, visibility, or culture?
Focus on one stage transition
Don't try to jump from Stage 2 to Stage 5. Focus entirely on reaching the next stage.
Set concrete milestones
Define what success looks like. Example: 'We're at Stage 3 when we run 8+ experiments per month across 3+ teams.'