geralt / Pixabay

We’ve all heard the analogy: data is the new crude oil. But we rarely focus on the “crude” aspect, which refers to a resource that, without refining, is useless. While data’s importance is widely recognized, few companies devote the time and resources to nurture and transform it into a monetizable asset.

I recently wrote about this concept in an article for Integration Developer News, where I discussed how challenging it is for many companies to make the leap from data collection and analysis to using that information to make a real difference in the bottom line. In order to make this transition, it’s essential that companies deliver on the value contained in their information and understand the steps required to tap into data as a new revenue stream or to reduce costs.

The most important step in effecting significant, organization-level change is that you must have a methodology in place that offers checks and balances and assures desired progress. To be effective, this methodology must also be easily understood and embraced among all stakeholders across the enterprise.

Borrowing key elements from manufacturing and other industries that convert raw materials into finished goods, we have defined an approach to transform data into a refined product. We call this rigorous, step-by-step process the “Data Value Chain.” There are five fundamentals to this approach—I’ll outline the first three in this post and expand upon the final steps in my next installment.

  • Data capture
    Of course, organizations must have the resources and ability to collect crude data. There are a multitude of advanced, quality methods to collect and store data from various relational and non-relational sources, as well as data streams—they may be more sophisticated than drills used in oil production, but the idea is the same. The key challenge of data capture is the variety and volume of data that can be captured. This poses a question about how to store this data, but more importantly how to process it for storage in such a way so that it is ready for analysis. Most companies throw the data into data lakes, but then it takes months for the users to get such data ready for analysis. The value of data is perishable; If you do not make it ready for analysis as soon as you capture it, the opportunities may disappear by the time it’s analyzes Big data presents actionable opportunities now!
  • Data quality and integration
    Once collected, data must be organized so that all the resources and components can be assembled into meaningful data units. Consider the various channels through which retail brands interact with customers: in-person, online, via social media, and so on. Given the various points of interaction with a single customer, there is a significant need for all the data to be integrated into a single view or golden record of that individual. Only then can accurate segmentation, cross-selling, and up-selling occur. Key in enabling this are proper data quality, governance and MDM (master data management), driven by the need for faster and more accurate decision making. Put plainly, incorrect data leads to wrong decisions. Ungoverned data is frequently unutilized or incorrectly utilized data. Therefore, it stands that not much value can be derived from the data assets without governance. We’ve been talking about “360-degree views” of the customer for years, but how can you get such a view without MDM—in other words, without having consistent records of your customers across all channels?
  • Data enrichment
    Frequently, organizations acquire external data and append it to their own records. Using the retail example, adding psychographic data—data about preferences and traits—is often useful for better understanding the customer. Combined with customer data, it provides marketers with a better picture of customers’ motivations, allowing them to develop more effective programs and offers.

These first three steps create a unit of data that is ready for analysis. The more complete the unit of data, the more insights can be derived from it. Data that is not used for analysis is waste, like a byproduct of refined oil.

Check back soon for my next post delving into the final two steps of the “Data Value Chain,” and actionable tips for companies to make the most of the resulting insights.