Fear has Replaced Apathy as the Number One Enemy of Data: Implications for Lovers of Data

By on

by Thomas C. Redman

Fear has replaced apathy as the number one enemy of all things data. That’s the headline result of my most recent scan of the space. The process by which I reached this conclusion found here. This fear reflects a growing sense that, sooner or later, data will penetrate every nook and cranny of every industry, company, and department, transforming the work, relationships and power structures. The fear extends beyond business people–even lovers of data feel it.

There are several implications for lovers of data (LoD). I include in this group data scientists, statisticians, data managers, data architects, data modelers, quality professionals, myself and all who labor to help our companies unleash the power of data. We didn’t do very well when the primary enemy was apathy. And fear is far more powerful, stiffening resolve, paralyzing ourselves and potential allies, and promoting irrational responses, even personal attacks.

This post outlines five mutually-reinforcing steps that we must take, individually, as a team within our companies, and as a network of professionals across companies, industries, and continents. Step one is to take a hard look in the mirror and recognize that we’ve failed our companies. This is a bitter pill, but consider:

  • We preach that data is a corporate asset, but we’ve not advanced the initiatives to make it so. In Data Driven, I proposed that data only qualifies as an asset if we take good care of it (quality and security), put it to work (make better decisions, create better products, innovate, etc) and advance management systems that recognize data’s special properties. Most data professionals support this thinking, but have not made it happen in their companies. Indeed, most companies can only be described as unfit for data.
  • We argue for business involvement, but have allowed ourselves to become buried in the bowels of IT.
  • We preach the importance of data quality, but focus the effort on cleaning up bad data, not preventing the errors in the first place.
  • We argue for the importance of meta-data, but have not put in place simple processes for keeping track of data sources and developing data definitions.
  • We’ve fostered an environment in which systems don’t talk to one another.
  • Despite data breaches being headline news, we’ve not put in place the needed multi-layer approaches to security.

This list can go on and on. The result, for the vast majority, is that our companies are not well-positioned for small data, never mind big data, advanced analytics, and data-driven cultures.   The look in the mirror must show that we LoDs have failed our companies by allowing this to happen. Further, and perhaps most importantly, we must not look to others to resolve the issues. Doing so is on us.

Step two is engage the business. LoDs talk about this all the time, then make only half-hearted efforts to actually do so. This needs to change. Start by developing a deep understanding of what the business actually does. Read the annual report, invite a counterpart for a drink after work (offer to pay!), meet for lunch outside the office (also pay!), and follow them around for a day or two. Observe what they do and, as they gain trust, ask them to explain what works and does not work from their perspectives.

Please note this is slow, patient work. Not everyone will make time for you, so don’t get discouraged when someone says, “Sorry, I can’t make lunch. Another time.”

Step three is to put the basics in place as quickly as possible. I count three items: High-quality, trusted data values; clear, easy-to-find and –understand data definitions; and solid security, as basics. This may appear a daunting task, particularly in light of growing data volumes, unstructured data, and all kinds of new sources. The key to cutting the task down to size is recognizing that relatively little data, perhaps ten percent, is truly important. Focus first on that data. And note that the only way to sort what data is most important is to engage the business.

Steps four and five fall to leaders, and those who wish to lead. In time, we must build better organizations for data. As I use the term “organization,” it includes people, structure, governance, and culture. This is a big subject, so here I’ll simply note that:

  1. We need far more data people,
  2. We need to get leadership for data out of IT,
  3. We need to get more data people in the line, helping others use data effectively, sort out the subtleties in what data really mean, improve quality, and develop a far deeper, more organic understanding of the business. I call these people “embedded data managers.”

Finally, step five is seize the initiative.  In the face of fear and in times of rapid change it may seem safer to adopt a wait-and-see, attitude, responding to events rather than attempting to steer them. Leaders must not let fear paralyze us.

For now is also a time of unprecedented opportunity: Opportunity to fundamentally alter the trajectories of our companies. To put data to work, to build more data into products and services, to empower everyone to make better, data-driven decisions. To capture data we’ve never captured before and, in so doing create entirely new industries. To innovate with data and advanced analytics, focusing not just on cutting cost, but on using data to drive growth. We’ve been talking about this for longer than I care to admit. We have the opportunity to make it happen.

Business managers at all levels want to know “what should we do about data?” A well-thought out plan, that aims high, but acknowledges the current state, sketches practical steps, and creates small wins, will receive a good listen. It will take courage to craft and present such plans and to engage in the thousands of conversations needed to build support.

Now is the time to seize the initiative.

Leave a Reply