Developing Relevant Metrics

Doug Hubbard says organisations measure what’s easy/cheap, not what’s useful. But how do we define what’s useful?

That depends on who wants to use it, and what they want to use it for. We need to build relevancy into our metrics reporting.

Metrics Pyramid

Over the past couple of months I’ve been working with a big national retailer. They’ve recently been looking to revamp their development, and ThoughtWorks were asked to assess how the retailer is doing and provide some recommendations. Among those recommendations was introducing the Metrics Pyramid.

Since the Lean Startup movement made metrics fashionable, there’s been a fascination with measuring everything. Unfortunately, a salt water lake isn’t very useful when you’re feeling thirsty.

The Metrics Pyramid is a structured way that can help you think about the metrics you want to track, by breaking it down by purpose/owner. It can help teams/companies think about which metrics are relevant (Who wants to use them for what?)

Metrics Pyramid.001

As you can see in the above, the steps of the pyramid have different layers of granularity. Those at the top, the high-level metrics, may be more relevant for the C-Suite, while the steps lower down, the interesting stuff, would be more relevant to the doers on the ground.

In (made up) practice

MCR is a Manchester based multi channel retailer. As with most multi channel retailers, MCR are seeing a trend towards online sales v. in-store. Jane is a Product Manager that works in the Search team for . Jane and her teammates ensure that search customers* find the product they’re looking for.

(*roughly 50% customers search, 50% browse. Which one are you?)

Jane’s Metrics Pyramid could look like this;

Metrics Pyramid.002.jpg.001

While the Head of may be more focused on Website Conversion, Jane herself is probably most focused on Search JTBD, and Search KPIs. Similarly, the User Experience Designer in the Search team is probably focused on Customer Insights. All the while, decisions are being made based on those metrics. If the metrics can’t drive decisions, they’re not relevant, and they need rethinking.


The Metrics Pyramid can give teams a structure to think about what the different level of metrics they want to track are, who they’re relevant to, and why. 

Once they have a decent understanding of what metrics are relevant, teams can then look at using Hypothesis Driven Development as a way to experiment with, and better, their product.

Questions for reader

  • How often do you see business/customer metrics?
  • Are they so hard to read that people don’t bother?
  • If people do read them, are they relevant enough to help drive decisions?

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s