In this session we played a little game using a pot several million pounds (fake, of course) and some features to decide whether and when to implement a re-usable features across several similar products.
Note: I don’t like the term re-usable very much. Either something is usable or not. If it is a usable API and I can inject system-specific dependencies it’s still not re-usable but more usable and well refactored.
In many ways, this is a variation on portfolio management and by adding the necessary input parameters (likely costs and revenue streams, uncertainties and probabilities) it becomes mainly a mathematical exercise to work out the numbers. In the end, you still need to make a proper decision which products you want to advance and that often requires some gut instinct. At least the exercise can weed out the worst choices.
Nice exercise but I don’t think I’ll took much home from it.