The Analytic Maslow Pyramid
During their digitalization processes corporations transit through different stages according to their analytic maturity, which is to say, their capability to transform data into convenience for their customers and employees, and value for their shareholders. Of course, the same can be applied to public administrations and the service they pay to citizens.
As in the well known Maslow Pyramid of human needs (that goes from basic items such as safety, food and shelter, to the most sophisticated social necessities), we can identify different layers regarding those business needs that can be leveraged by data and analytics.
Layer 1: In the early stages of digital transformation very basic issues must be solved. For example: every interaction with the same customer should be recorded with a unique customer ID, regardless which is the channel used by the customer. Phone calls, on-site and on-line interactions should be linked in order to describe the journey of a client. Of course, data should be saved with enough granularity to allow historical track of those interactions, a trade-off between storage costs and future applications of the data must be reached.
Layer 2: Your customers are linked between one another and with external agents, graph analysis and clustering techniques are powerful tools to make visible those hidden relations in your data. Besides, they are the foundations of good recommendation algorithms, a key piece to take the right step from traditional marketing to relevant value offering.
Layer 3: To provide a positive user experience your services must anticipate your customers’ needs. Without becoming intrusive, you have to predict when and how they are going to require each or your products: convenient proposals on time are among the most amazing things one can expect from a service provider.
Layer 4 and top of the pyramid: Reality is multi-faceted, but too often we only see one side of it. Most companies are overwhelmed with the huge and growing amount of data they gather from their clients. They are desperate to transform that raw material into tangible value with scarce human resources, as Data Scientist are disputed profiles. That is the reason why they are commonly lost in an introspective focus, analyzing only those direct interactions that they have from their customers, and forgetting about the rest. But, without context, information is biased. Corporations must provide transparent incentives to their customers that make them share data to enable a complete 360º vision of their preferences. Yet, beyond personal data there are multiple external data sources that put things into context and provide meaningful insights. One of my favorite is spatial information: the perfect frame for a fact is where and when it took place.
This easy scheme tells us about an evolutive path in data and analytics applications. It is perfectly understandable that every institution has to address basic issues described in layer 1 before climbing in the pyramid towards more sophisticated goals. But having primary questions unsolved should not prevent us from thinking big and being ambitious. The other way around, we must not forget that, after all, data are always just digital shadows of a much more complex reality.
