How to design a metric that matters

Measuring the right things helps us make better decisions.

Here is a simple four-step process that you can use to design metrics that matters the most to you:

Describe the intent. What is the purpose of this metric? Who is it for? What is it connected to, if anything? Think about KPIs, objectives, targets, etc. What are the desired changes in outcomes and behaviours?
Describe the metric. What are you going to measure? How will you measure it? Is it an actionable metric? Is it understandable?
Test the metric. What is the easiest, quickest, and cheapest way to introduce this metric? Can you do it with a smaller number of employees to test the metric before your roll it out to the whole organisation? Have you noticed intended changes after measuring it for a relevant period of time?
Continuously review and adjust the metric. Does the measure create intended changes, which in turn positively influence the business? How can you communicate it better? How can you streamline the measurement process and reduce the error margin?

If above feels daunting, let me share Douglas W. Hubbard's four useful assumptions when it comes to establishing new metrics:

It has been measured before.
You have far more data than you think.
You need far less data than you think.
Useful, new observations are more accessible than you think.

You aren't designing metrics in a nuclear power plant, but for your business. The worst thing that can happen is that you waste some paper.

On the other hand, the upside is massive. Any metric that has casual relationship with revenue, profitability, or any impact that really matters to you, is another lever that you can use to drive them.

That's how you move from a black-box hit-and-miss mindset towards a genuine evidence-based and data-driven business where you can create repeatable success.