The unique reusable nature of data and information when combined with digitization of every aspect of enterprise data makes for a very potent mixture for leverage. As data-driven professionals, the question we need to ask ourselves is: are we leveraging this enterprise data?
First, let us look at what would this leverage look like:
- Answers for previously unanswered questions and also better answers: The first key leverage is getting answers to questions that we previously thought were impossible to answer. We can also find ways to ensure that these insights can be achieved faster and cheaper and made more granular.
- Questions that you may have never thought about: If data were to speak to us and provide, say, 10 patterns, without even understanding what it is serving up, it is both conceivable and already seen in practice that while a part of the predicted patterns would be already known to the business, another part may end up making no sense in real-world (though it did so to a machine), but there would be a part that would be most interesting. These last set of patterns would raise unseen questions and lead us to the second key leverage — coming up with questions we never thought of having.
- Treating data as an asset: As data assets are very tangible, but are not governed by the same commercial laws (balance sheet, taxation, etc.), they are not subject to the same rules of financial transactions. Therefore, rules of bartering data would be different than bartering any other asset like buildings, machine, bonds, equity, or any other financial tool. There are a huge number of ways in which it can be leveraged without incurring overheads associated with a typical financial transaction.
Now you need to ask yourself if you are maximizing the advantage you can get out of your rich data assets?
If the above outcomes (or their variations) were to be your goals for your enterprise, then the enterprise data needs to be made available per the following three simple guiding principles:
- Discoverable: Synergy requires two or more entities to interact. If they react collaboratively, the result may be 1+1 = 3. If they do not, the result can be detrimental. But, either way, the basic requirement is interaction. For interaction, we need to be able to see each other. So, for data sets to act synergistically, they need to be seen. Consumers of this data (people or machine) should be able to easily discover, identify, and understand the data and their attributes.
- Joinable: This is a logical extension of the point above. Just putting two things next to each other, although discovered, will not lead to interaction unless they have something they can find in common. We need to allow more fuzzy joining of the data beyond the standard id to id and name to name type of equal joins. So, through the use of heuristic matching or triangulation, we can allow consumers to join data that were not considered on the same plane before.
- Governed: Of course, data requires governance. From compliance to the risk, and even just for core measurements of performance, we need to monitor who, how, and when the data is accessed. The core challenge here is the scale of governance and knowing if such a scale would make agility/innovation near impossible as it takes just too long to try something new. Governance will need to provide security and compliance, but we will need to do it in a way that provides agility to the consumer and not overly constrain them.
The fundamentals of gaining a competitive edge still remain the same, and one of these fundamentals is to leverage an asset to maximize market advantage. Enterprise data is massive in nature, and the information derived from the data can offer a huge business advantage to any organization, provided it’s extracted quickly and accurately, and the data sets used are authentic and harmonized.