Why I Think the DIKW Pyramid Should not Be a Pyramid

The world of big data is evolving rapidly and the general public is starting to adopt it as their own. The question that often arises is; we want to do ‘something’ with big data, but what? For large companies this question is easily solved by hiring data scientists. Sure, there are many useful tools out there that show graphics from which you can extract conclusions. However, small and medium-sized companies often don’t know where to start putting big data into practice.

At Datatrics we believe in doing this a bit differently. We developed a platform for small and medium-sized marketing teams that turns big data into concrete actions, so they immediately know how to approach their audience.

The foundation of our platform is perhaps even more interesting. We used the DIKW Pyramid to turn the big data our platform collects, into concrete actions for marketeers.

However, I don’t believe the DIKW Pyramid should be a pyramid at all. Turning data into actions does not consist of filtering and making the amount of data smaller. It consists of transforming the data and perhaps even gaining more data in the process of learning and gaining wisdom from the data. Therefore, I believe the DIKW Pyramid should in fact be shaped like an hourglass. So let me elaborate on how such an hourglass shape would work for our platform.

Data

The first step is collecting data. This data needs a context, otherwise it cannot be used in the process of turning data into wisdom. For example, the number 34 alone doesn’t have any context. However, if it’s found in an ERP-system as the amount of office chairs sold the past month, it can be used in the DIKW process.

Information

However, just having data in different places, although within a context, is still not very useful. You want to be able to turn the data into information. By combining the data, different correlations can be found within different sources. As a marketeer, this will help you to understand the market; identifying what is relevant.

Knowledge

After identifying the data and turning it into information, you can transform it into knowledge. Knowledge being a refinement of information because that’s how you extract value from data. A data scientist will be able to do this ‘manually’. However, smart algorithms can also do that for you. They can find correlations within the data and make the information more actionable, transforming knowledge into instructions you can easily understand and follow.

Wisdom

But it doesn’t stop there. You can and should learn from the actions you perform. Measuring the performance of your actions results in more valuable insights. And even better, using measured performance based on past actions as input, will result in more accurate predictive models that make suggested actions more relevant.

This feedback loop refines the knowledge you previously gained, and thus gives you wisdom to act upon in future interactions. By increasing your wisdom through measuring the performance of your actions, new data will be generated. This extra data needed to gain wisdom, is why I believe the DIKW pyramid should be shaped like an hourglass.

Love to hear your thoughts, comments or experience on this!