Listen to this article:
Imagine it’s all done. Well, it’s never all done, but imagine it’s all there: definitions, expected content, data models, profiling, documentation, support processes. Everything we’ve spent so much time trying to make happen. It’s all there.
What now?
Take a minute and try to imagine it.
What do you see? Before continuing, put your thoughts into a comment or send an email.
Through years of limited success at best, we’ve trained ourselves to not even consider the question. Furthermore, too many of us in the data and analytics business have become comfortable with the no progress. Yes, it gets frustrating at times. We try, we succeed enough to claim we’ve accomplished something. We move on or keep trying. There’s no shortage of companies hiring consultants to do data governance, information management, and analytics. And there’s no shortage of experts ready to prescribe what’s worked for them in the past (whatever worked happened to be).
But the result has been decades of the same arguments, the same artifacts, the same templates, the same assessments, the same methodologies, the same resistance, and the same glacial progress. We put together plans and presentations. We create data models. We survey users. We create councils and review boards. We hold meetings. Leadership thanks us for our efforts. Not much happens. The cycle repeats.
It’s easy, and understandable, to resist imagining. We’d be lucky to convince them to do the basic stuff (whoever them happens to be). It would be a miracle to even get that far. If we did, we’d be like the dog that caught the car or the teenage boy after the girl said yes to a date.
The movie Cast Away came out while I was working at FedEx. Tom Hanks played a systems analyst who survived a plane crash and washed ashore on a deserted island. Throughout the film he tries to leave the island on a raft, but is always forced back by powerful waves breaking over the surrounding reef.
Sound familiar?
In one last desperate attempt to escape, he rigs a sail that deploys just as he reaches the barrier. The force of the wind pushes the raft past the breakers and into open water away from the island.
For us, AI is that wind that can push us past the breakers.
Of course, escaping the island was just the first step in the long journey home. So, too, with us.
Now is the time to think about what comes next: Data Governance’s Second Act.
The future won’t happen immediately, but it is coming fast. AI is both the accelerator and the motivation.
The foundation of this imagined future is applying AI to the activities encompassed in the DMBOK wheel so that they are actually done. AI will generate the metadata, data models, DDL, queries, analyses, dashboards, communications, workflows, and more. This allows data governance and information management capabilities to achieve critical mass where they become standard operating procedure. After all, this work never stops. There will always be new applications, new data elements, and new content.
With AI facilitating and accelerating the work that we’ve been so comfortable doing for so long, effort and attention can be directed elsewhere.
Our focus needs to shift from creating artifacts to creating value.
I spent time brainstorming three categories of what could come next: facilitating the work of data consumers, expanding our understanding of the data, and imagining novel uses for data. The first two are inwardly focused, improving upon existing capabilities. The third is outwardly focused, enabling new activities.
Facilitating the work of data consumers:
- Automated Insight Discovery
- Metric Efficacy Optimization
- Data Products / Data Mesh / Data Fabric
- Data Estate Optimization
Expanding our understanding of the data:
- Unstructured Data Content Quality Assessment (this is a big one that’s barely been touched and can be greatly facilitated through AI)
- Automated Ontology Discovery
- Infonomics for Real
Imagining novel uses for data:
- Self-Healing Operational and Analytical Systems
- Automated Business Capability Implementation
- Organizational Introspection, Learning, and Improvement
- Data Monetization (always seems to be on the list but rarely done)
I plan to go into more detail in future articles, but these should get the conversation started.
Most organizations don’t have a sold enough foundation to pursue most of these right now. I mean, you can try, but the risks are great. This transition period gives us a little time to get our act together.
In five years … no, three years … data governance, information management, and analytics will look different and be practiced differently than today. We’ll wonder how we tolerated riding the treadmill for so long. The transition will be uncomfortable, and maybe a little scary, but it’s better to prepare now than to be surprised and have to play catch-up later (I’m looking at you, AI Readiness).
