Listen to this article:
Imagine you’re a product manager in Marketing. You use data every day. Then someone tells you, “Congratulations, you’re a Data Steward.”
Your first question is perfectly reasonable: What am I expected to do differently tomorrow morning? Data Literacy training, of course. Traditional Data Literacy training begins with artifacts: business glossaries, metadata, stewardship processes, data quality rules, lineage, governance councils, and more. Participation often feels burdensome, as though it’s expected merely “for its own good.”
Furthermore, those investing time creating Data Governance assets and those benefitting from those assets are often different. It appears to be a significant and demotivating disconnect. The Data Accountability Chain shows that they are actually connected parts of the same accountability system.
Information Management artifacts are not the objective.
They’re accountability assets that enable different people to answer different accountability questions. Everybody participates in Information Management not because of Data Stewardship or slogans or policy, but because everyone participates in the Data Accountability Chain.
Last week we looked at how accountability works. Now we’re going to reinterpret Information Management through that lens, starting with the role of the Enterprise Data team.
Enterprise Data teams have traditionally been viewed as the centralized group of technical professionals responsible for managing, storing, and organizing the company’s data assets. Recently, the positioning has changed, treating data as a strategic product that empowers business teams to make decentralized, autonomous decisions. This is a very good trend. In fact, I have extended that decentralization within the context of Data Products to include development teams.
Now I’d like to go even farther in reorienting the Enterprise Data team:
The purpose of the Enterprise Data team is to ensure that important organizational decisions can be explained with credible data.
This isn’t just Data Products, or empowerment, or decentralized autonomous decisions. The Enterprise Data team maintains the organizational capabilities and infrastructure that connect executive confidence to operational evidence. Metadata, lineage, data quality, governance, stewardship, Data Contracts are not objectives in themselves. They’re the implementation of organizational capabilities that enable accountable decisions and allow an executive to say:
Yes, I trust this information because we have a system that continuously measures, monitors, and improves it.
The Enterprise Data team supports the entire Data Accountability Chain without owning any of the individual links. It may help create business definitions, assign data owners, or document lineage, but its ultimate accountability is to create the processes, standards, and infrastructure that make those activities possible. Increasingly, AI is performing much of that work itself.
This brings us to the most important question.
What is Information management for?
We have been answering this question from the wrong perspective for decades.
It is the reason that data professionals have been missing the mark when speaking with leadership (and the business and development). No wonder we get so little traction. No wonder we get so little support. We talk with leadership (and the business and development) about artifacts. They couldn’t care less. To be clear:
Information Management does not exist to produce artifacts, it exists to enable accountable decision making.
Read that last sentence again. Take a minute to think about what that means before continuing.
Having data models, lineage, access rules, definitions, and all that stuff is not the objective. Having reliable, accountable decisions is the objective.
This is a new way to understand our entire profession.
Let’s revisit the Information Management artifacts we’ve all relied upon for years through the lens of accountability, reframing them as organizational capabilities that enable accountable decisions. You may notice that the artifacts are nouns but their purposes are verbs. This shift is important. Stop defining Information Management by the artifacts it produces and start defining it by what those artifacts enable.
| Information Management Artifacts | What It Enables |
|---|---|
| Business Glossary | Shared Understanding |
| Metadata | Explanation |
| Data Catalog | Discovery |
| Lineage | Traceability |
| Data Quality | Confidence |
| Data Products | Reliable Use |
| AI Documentation | Explainable AI |
| Stewardship | Ownership |
| Metrics and KPIs | Learning and Improvement |
People outside the data profession rarely care about the artifacts themselves. They care about whether they can trust the number they’re using to make a decision. So…
Instead of asking, “Do we have a data catalog?” ask, “Can people find the data they need?”
Instead of asking, “Do we have data quality?” ask, “Can people confidently use this data?”
Each artifact requested and each artifact produced can now be judged by a single question:
Does this artifact (and its content) help somebody to answer “Why do you trust this?” Does its absence diminish trust?
In the context of Data Accountability, these artifacts exist to provide evidence and assurance.
For decades, Information Management has defined artifacts. Accountability explains why they matter.
AI changes the economics of Information Management, increasingly performing the activities and creating the artifacts, but it does not change the organization’s obligation to answer the “Why?” questions. In fact, it makes that obligation more important.
AI commoditizes the “how” but it doesn’t eliminate the “why.”
“Fine. But if accountability is so important, isn’t this really a management or culture issue rather than an Information Management issue?”
Yes, and that’s the point.
Information management cannot create accountability, but it can create the information, evidence, assurance, and transparency that make accountability possible.
Accountability, especially when it comes to data, is an organizational obligation. Information Management is one of the mechanisms that enables an organization to fulfill it.
I would be willing to bet that very few organizations actually think about data from that perspective. I’d be willing to bet that very few organizations ask the question, “Why do I trust this data?” I’d also be willing to bet that when asked, the answers are trite, perfunctory, and show little to no depth. “It’s the way we’ve always done it,” or, “We have a talented team.” Would you treat your quarterly financial data that way? Would you report financial figures from unidentified sources containing calculations nobody understands? Of course not, yet we routinely accept comparable conditions for the information that drives operational and strategic decisions.
Here’s a recommendation to get you started: replace artifact-oriented maturity models with accountability-oriented models. When you ask the questions in that way, see if you don’t get more support and interest from leadership, the business, and development. Just because they aren’t interested in seeing the risks associated with poorly understood data doesn’t mean they’re not there.
Next week, I’ll continue this transition from artifact to capability-based Information Management by taking a close look at the DMBOK Wheel. Maybe it’s time to rethink its purpose, especially when we consider how little visibility (rightly so) and how little interest (unfortunately so) it has generated for Information Management beyond us data professionals.
