In A Nutshell

We deliver features in our products and services because we believe they will create value for our customers. That value in turn contributes to the success of our organisation. Before we deliver features and services, what we have is a belief - in other words, we are betting on a successful outcome. After we deliver features we should gather information to confirm that our belief is correct or incorrect.

The information allows us to decide how to act. If the data confirms we made the right bet we can continue to formulate our bets in the same way. We can decide to pursue the bet further in forthcoming periods of work. Alternatively, if the data demonstrates that we made the wrong bet then we can reduce the priority of the investment. In an extreme case we can decide to stop the investment completely.

To get the right information, we need to define how we are going to measure value. Unfortunately, this can be complex to achieve effectively. Value is realised outside of our organisation and delivered into it. This means that value can be impacted by many other exogenous factors, not just the change or feature we have introduced. Disentangling these multiple, potentially conflicting influences is very challenging.

To help simplify the tangle of influences, we often use a combination of direct and indirect measures. Direct measures are immediate or proximate measures of the value we are betting on. Indirect measures examine influences that we believe are correlated (or anti-correlated) with the value we are betting on. If we see signals from direct measures and correctly correlated signals from our indirect measures, we have more confidence that at least some of the change in our value metric is related to the change that we have made.

Implementing Practices

  • Sustain Metric Definitions

  • Analyse Metric Data

  • Use Data To Decide