Experiment Quality Score
Get an at-a-glance quality assessment of your experiment design with AI analysis that scores hypothesis clarity, metric validity, testability, and specificity.
What is the Quality Score?
The Experiment Quality Score is a collapsible widget on experiment detail views that provides an overall percentage score (0–100%) based on four dimensions of experiment design quality.
Quality Dimensions
| Dimension | What It Measures |
|---|---|
| Hypothesis Clarity | How clear and specific is the hypothesis statement? |
| Metric Validity | Is the target metric appropriate and measurable? |
| Testability | Can this experiment be properly tested within constraints? |
| Specificity | Are the implementation details specific enough to execute? |
Each dimension is scored independently and displayed in a 2×2 grid with progress bars.
Score Levels
| Score | Label | Icon | Meaning |
|---|---|---|---|
| 80–100% | Excellent | ✅ | Ready to launch |
| 60–79% | Good | ✅ | Minor improvements possible |
| 40–59% | Fair | ⚠️ | Should refine before launching |
| 0–39% | Needs Work | ❌ | Significant issues to address |
How to Use
- Open any experiment's detail view
- Find the Experiment Quality collapsible section
- If not yet analyzed, click "Analyze Quality"
- The AI calls the hypothesis refinement engine to score your experiment
- Review the overall score, per-dimension breakdown, and AI suggestions
AI Suggestions
When the quality analysis detects issues, it provides specific, actionable suggestions displayed as bullet points. These may include:
- Missing audience segment definition
- Vague success criteria
- Unmeasurable outcomes
- Missing rationale or causal reasoning
💡 Tip: Aim for a quality score of 80% or higher before investing resources in running the experiment. Use the Hypothesis Refiner and Metric Validator to address issues flagged by the quality analysis.
