Testing for Actionability
You'll learn how to validate that a metric actually tells you what to do when it changes.
The Actionability Test
A metric is only valuable if it guides action. The actionability test is simple but ruthless: If this metric moves up or down, do I know what specific action to take? If the answer is "no" or "maybe," that metric fails the test.
This is where many seemingly impressive metrics collapse. Consider "brand awareness" increasing by 5%. What do you do? Keep doing exactly what you're doing? Try something different? The metric moved, but you're left guessing.
How to Apply the Test
Ask yourself three questions:
- If the metric goes up, what would I do? (Continue? Accelerate? Investigate?)
- If the metric goes down, what would I do? (Fix something specific? Change strategy? Which lever would I pull?)
- Can multiple people agree on the action? (Or does everyone interpret it differently?)
If you can't answer these clearly, the metric isn't actionable.
Real-World Example
Compare two metrics for an e-commerce site:
- "Total site visitors" — Goes up 10%. What now? You don't know if it's bots, wrong audience, or valuable traffic. Not actionable.
- "Checkout abandonment rate" — Goes up 10%. Action: Investigate checkout friction, test simplified forms, check payment processor issues. Highly actionable.
Notice the difference? The second metric immediately points you toward specific product areas and experiments.
Why This Matters
Vanity metrics often fail the actionability test. So do overly aggregated metrics or those disconnected from your control levers. Testing for actionability forces you to design metrics that drive decisions, not just dashboards.
Key Takeaway: A good metric doesn't just measure—it tells you what to do next. If movement in your metric leaves you shrugging, find a better metric.