One excellent way to make different individuals understand what is meant by "all the variables needed" for generating knowledge or data-based decision-making is to reflect on it through the Job-to-be-Done theory developed by Clayton M. Christensen. While the theoretical framework is designed to address innovation and new market openings, it also gives us a good perspective into the development of data products and services.
Job-to-be-Done theory is about understanding what a client is trying to accomplish and why. One frequently used example of this is a situation where a man in his 40s walks in to a newsstand and here in Scandinavia he would probably buy some Sports Magazine, car magazine or afternoon newspaper but he ends up buying some baby magazine. Why? Well, because his wife had asked me to buy this magazine when she came home from work, with interesting articles about the upcoming family addition. Your data does not explain this.
So how does this example relate to the data economy or data products? Imagine yourself as the owner of that newsstand that mentioned earlier and the newsstand would be an online store. When a potential user logs in, you start to recommend and sell different products based on age, gender, and perhaps time of day. Profiling and recommendations go wrong. Sure, much more civilized algorithms for recommendations and sales on digital channels are already available today, but the lesson of the story lies in how these customer needs should be thought of.
Christensen has found that when getting to know and trying to understand a customer all the time, more and more people go in the wrong direction in sales and successful service planning. There should be more focus on thinking and investing in what the client is really trying to accomplish and why, i.e. Job-to-be-Done.
In the data world, it is easy to understand that if in a situation like this one thinks about the sources of information from which data can be collected about a person’s age, gender and time of business, it is very easy to sell the wrong things to the wrong people and at the wrong time. But if information about a possible family addition were also added as a data source in this example, there would already be a reasonable explanation for the purchase and one more recommended item. This easily makes sense with such a simple example and this is the work that every data team should do; what knowledge we have to model things now, what we may be missing, and what factors explain these anomalous situations. In this way, you will constantly gain an understanding of the things you really want to know and that will increase your competitiveness.
So, thinking what data you need in order to understand all possible scenarios in more depth, you will end-up to generate understanding that no-one else has. With this kind of data products (answers to Job-to-be-Done questions), you can easily capture more business value.