Almost everyone in my IT Odyssey interviews expresses certainty that their enterprise finally views data as a corporate asset. I could make a lot of money selling bumper stickers and executive designer T-shirts emblazoned with that catchphrase.

But when I ask for the detailed, resourced specifics of what exactly is being done to actively manage data as an asset, most said:

  • We’re hiring data scientists and beginning to apply data analytics
  • Our progress has been slowed by the terrible state of our data

Neither quite gets to the heart of the matter.

My older son is a well-respected, high-end professional athletic trainer. For 30 years, he has drilled it into me that no matter what I spend on exercise equipment, it’s all worthless unless I first do the work of strengthening my core.

And that’s what’s missing from most of the analytics initiatives I’m hearing about: a strong data core. Unfortunately, many enterprises haven’t realized that no matter what they do to get their IT and analytics functioning effectively, all of it will be worthless without doing the work on their data core. They have decades of accumulated, redundant, conflicting, incompatible data and are paying the price.

Everything But the Core

Here are just a few examples of what I hear. A financial services CIO recently told me her organization has experienced the futility of reorganizing its massive, disorganized data warehouses (once again) into data lakes, hiring more data scientists and buying a new analytical tool.

A medical industry CIO with zillions of petabytes of legacy accumulated data from over 40 global acquisitions told me that when it comes to master data management, MDM stands for “major data mess.” He proudly mentioned he’d assigned two full-time analysts to the clean-up and rationalization tasks. I thought he was kidding me. Two analysts are not going to do much to sort out the company’s data management needs.

A data executive of a distribution company agreed that many of his organization’s data issues were the result of shotgun mergers and technology changes. However, much of the data mayhem was made worse by years of poor data hygiene. For instance, while they assigned data tasks, they failed to provide guidance and enforcement for these responsibilities, which included:

  • The producer/data originators were responsible for ensuring the data was consistent, not already available elsewhere and accurate as it left the originating data source
  • The owner of the data factory was tasked with making sure the data was accurately placed in locations that were well defined and easily accessible for the appropriate consumers  
  • The data consumers were responsible for making sure the information was used as defined and from the appropriate source

His company never seemed to have the time, resources or discipline to do data management correctly, he lamented. Yet he knew that effective data management would lead to a healthy data core with the following characteristics:

  • A relatively precise and up-to-date differentiation and specification of the core data elements
  • A minimum amount of redundancy and inconsistency across the enterprise and technologies for the core elements
  • Monitoring and maintaining of the elements
  • Clear ownership, stewardship, rules and governance for the data elements
  • The ability for the data elements to be easily and readably known, available and accessible to users and developers
  • Hosting of the data elements on current technologies

Data as a Business Enabler 

It’s not an overstatement to say that data architecture and management have been the neglected disciplines of the past few decades. Ten years ago, while embracing the four TOGAF domains, most enterprise architectures focused on technology. More recently, the focus has been on business architecture comprised of capabilities and value streams. Efforts today such as low- or no-code, citizen and/or Agile development, cloud migration and data analytics often pay the price for this core data neglect.

IT executives who believe they have a healthy data core often suggest the following 11 essential steps and capabilities for creating one:

  1. Major dedicated resources, as well as broad, coordinated and funded proactive efforts across the business and IT over several years
  2. A real data architecture function to guide core data efforts
  3. Periodic data health checks that include agreement on critical, core data; an inventory of core data; and a health profile of that core, critical data
  4. A data governance board to approve data architectures and adjudicate, reconcile and monitor data remediation plans
  5. Data ownership, as well as education, on how to be a conscientious data owner
  6. Institutionalization (through education, promotion and mentoring) of data hygiene disciplines and practices as described above by the aforementioned data executive
  7. Data remediation priorities based on health, critical need, ongoing projects and financial considerations
  8. Specific data remediation plans, programs and projects
  9. A real data management function and process
  10. Proactive IT efforts to introduce new data guidance and management
  11. A twice annual review of data health and remediation by the CIO and a senior business leader

This may sound like a lot of resources and bureaucracy, but assets need to be managed if their value is to be realized.  Senior business and IT executives should start by scoring how well their enterprise does on the 11 steps above.

A healthy, digital information technology backbone can only add value if it resides on a healthy data core. And once you’ve got your data core in place, let me know – I’ll put you on the waiting list for a T-shirt and bumper sticker.  

Bruce J. Rogow

Bruce J. Rogow is a Principal at IT Odyssey and Advisory in Marblehead, Mass. Known as the counselor to CIOs and CEOs... Read more