In A Nutshell

As a team we play the role of trusted advisor to our customers. In the technical domain we can play this role well. In the business domain we will normally have less knowledge than our customers. How can we play the role of trusted advisor when the customer knows more than us?

Even with their greater experience, customers will often lack awareness of how specific needs can be realised technically. Or there will be unresolved details for specific needs. Sometimes we will even find conflicting needs and will need to help the customer to resolve the conflict. In our role of trusted advisor we provide the mechanisms by which incompletely stated needs can be refined and conflicts can be resolved.

We need to gain data about which needs and implementations of needs will be most valued by our customers. We need to obtain this data within the constraints of the technology choices we have made. Experimentation is our main practice for elaborating and resolving our understanding of customer needs.

Very often we will need to operate a series of activities. Each activity will feature one or more experiments to help prove the customers’ needs. Our activities are sequenced so that we test what we believe are the most viable hypotheses first. We also agree how we can exit from the sequence of activities early. Which experiments provide sufficient evidence so that we can abandon lower priority experiments at an acceptable level of risk?

Many of the experiments will involve A/B testing. These experiments allow us to evaluate two or more alternative implementations at the same time. The result of such experiments demonstrates which alternative (or, sometimes, combination of alternatives) provides the best value for our customers.


Implementing Practices