Decades ago, when I entered the industry, I joined a department called Data Systems. Many of my wet-behind-the-ears colleagues worked at other companies in their Management Information Systems (MIS) departments. We learned how to make the big iron mainframes crunch numbers and deliver reports to the line units.
Sometime later, (the 1990s if I recall correctly), with the widespread adoption of PCs and Local Area Networks, it became commonplace for those same departments to be renamed Information Technology (IT). In the intervening time, our industry-wide love affair with technology (the ”T” in IT) has continued unabated. Sadly, I cannot say the same for the other letter in IT; “I” or Information. Information seems to have taken a backseat to its sexier partner.
As I encounter legacy data sets in the various jobs I perform, I’m often disappointed in their level of fidelity and completeness. It seems that the developers and integrators who put the legacy systems in place cared more about the software than the data that was being accumulated. This makes me sad. But it doesn’t surprise me.
You see, there is no one-size-fits-all solution for creating high quality data. It takes hard work. It takes understanding the customer’s business. It takes time, and patience, and you must observe how people actually use the system(s). It takes building the right balance of incentives and consequences. And it takes getting the right people (management) to consume the information coming from the systems. It takes leadership from management to convey the importance of high quality, timely, and relevant information to their organizations.
You cannot put this process in a box and shrink-wrap it. And that’s why the computer industry favors technology over information. They can put a shiny piece of technology in a box and make a tidy profit. But they just cannot deliver high-quality data. You have to do that part yourself.
I submit to you that the only thing of lasting value that we (IT professionals) create is high-quality data. Everything else is a distraction. It’s the Data, Dummy.