Most software products have multiple opportunities for using predictive analytics to enhance decision-making. Because the use of predictive analytics is new for many software companies, and because handling decisions explicitly is also new, care must be taken not to bite off too much at once. A land and expand strategy that begins with a localized effort and expands systematically is going to be the most successful.
As you look at predictive analytics and try to decide where to focus, remember to keep the decision in mind. The value to your customers will come not from a predictive analytic model, but from the better decisions that can be made thanks to it. Keeping the focus on decisions ensures that you can articulate the value of each predictive analytic model in business terms. Over time you can add more decisions, reusing predictive analytic models where possible and generating new predictive analytic models where needed. Each decision adds value, each use case for predictive analytics is described clearly.
As this decision-by-decision expansion makes clear, adding predictive analytics to a software product is not a one-time, one-release proposition. Most products have multiple decisions that can be improved and multiple predictive analytic models that can be built. In addition customers will become more sophisticated once they see the power of predictive analytics. As each new decision is managed or supported, each new predictive analytic model developed you should be developing an in house competency around predictive analytics. It does not need to be a huge group as the use of automation can help a small team scale and is anyway essential for a software product.
Your objective should be to build a competency that is broad, able to see how predictive analytics could be used widely within your software product(s), rather than one that is deep in a single area. Plus you want to empower your customers to build their own analytic competency. Finally remember that predictive analytic models age, they degrade over time. It is important that your software recognize this and build in capabilities to monitor the predictive power of the models in use, learn which ones are working well and which ones need to be refreshed.
The next generation of software is going to analytic. It’s not just going to report on its data, show it on dashboards, it is going to use it to analytically drive more effective, more precise, more profitable behavior. Software companies face a harsh choice—either figure out how to use predictive analytics to power better decisions in your software or fall behind competitors that do. Begin now.