Knowledge bank

In all problem solving, it is important to understand the problem before trying to solve it. So that a solution is not built for the wrong problem.

Once the problem is understood, it is time to transfer reality to a model. For the data points to be different outcomes in the model and not just different KPIs without significance.

When it comes to machine learning, this is even more important when using data to understand and solve the problem, then you must also quantify the problem. If an understanding of a customer journey is to be created, it can be good to ask the following questions to a sales manager, for example:

Who are your customers? Why do they buy your product? Are there any buy signals before a purchase is made? What distinguishes a good customer from a less good customer? How much do we lose if we do not actively market our product?

It is optimal to get answers to such questions and at the same time have access to data. Then it is possible to more easily understand the problem and thus build a solution for the right problem.