Viewpoints

Enterprise View

Jill Dyché, partner in Baseline Consulting


Jill Dyché, a partner in Baseline Consulting, says there’s no template for data governance; your program will depend on your organization.

Data governance
takes off

Focus shifts from back room
to boardroom.

As executives become more serious about managing data as an asset, data governance has gone from the back room to the boardroom. Teradata Magazine recently sat down with Jill Dyché, a partner in Baseline Consulting, to discuss how data governance is evolving.

 

To start, how do you define data governance when you’re explaining it to your clients?

Well, we have the requisite formal definition that includes processes and frameworks and decision rights. But the bottom line is that data governance is the business-driven oversight of enterprise information.

Sounds simple enough, but is it?

No, it’s not. In fact, people have underestimated its complexity.

You frequently speak about the many different dimensions of data governance. Why so much variety?

When Kimberly Nevala and I were first developing our TDWI [The Data Warehousing Institute] course, we thought, “How are we going to do a whole day on data governance?” Now we have a hard time keeping the class to a day, and we’ve had to reduce our material so attendees can get to dinner! Seriously, you really have to deconstruct data gover­nance into its components in order to do it justice.

And what are some of those components?

There are really four main areas of focus:

  • The business framework is the creation of a data governance charter, guiding principles and the business drivers. It’s essentially making the case for data governance and design­ing how the business will oversee information.
  • Processes and policies is where we actually break down the decisions around corporate data. There may be more than one process for this, with mul­tiple layers of stakeholders.
  • Data management is the tactical execution of the business poli­cies. This is where IT engages.
  • The execution processes is when the council convenes, agendas are drafted and fol­lowed, and automation comes in to manage workflows and enable data stewardship.

The formation of a data governance council typically gets a lot of attention. Why isn’t that on your list?

Good catch, but that’s intentional. We find that many of our clients have made the mistake of forming a council too early. I call this the “Kickoff and Cold Cuts” syndrome of data governance: Someone calls a lunch meeting of data stakehold­ers, brings in lunch from a local deli, and kicks off data governance.

All of the invited stakeholders attend the meeting because, after all, data’s important—and because they want to snag one of those ciabatta sandwiches. They all talk about why data should be managed as an asset, and there’s universal consensus that something needs to be done. And then they schedule a follow-up meeting, which doesn’t draw as many people.

"The bottom line is that data governance is the business-driven oversight of enterprise information."

And there’s no free lunch!

That’s the punch line! But you get the larger point: The focus is on the “who” before the “what” and the “how.” The moral of the story is you shouldn’t prematurely convene a group of people around data gover­nance before you’ve designed the framework and processes.

So without a framework and process in place, it falls apart?

Absolutely. And sometimes those teams don’t get another chance to resuscitate it. It’s easy enough to call a meeting and get everyone to agree that something needs to be done. It’s much harder to design a sustainable program.

You’ve used the word “design” several times. Can you explain that?

It’s another thing very few companies have done. It might sound obvi­ous but there’s no template for data governance. It looks different at every company. That’s because corporate systems, cultures, business problems and incumbent organizational structures are so different. If you’re launching data governance, you need to be able to answer the question: “What will it look like here?” And if you can answer that before you’ve engaged stakeholders, you’re much more likely to keep their attention.

At the Teradata PARTNERS Conference, you spoke about launching data governance via business intelligence [BI] pro­grams, but your book [co-written with business partner Evan Levy] advocates attaching data governance to master data management [MDM] initiatives. Which approach is best?

It depends. We’ve seen a lot of BI programs work perfectly well with­out data governance. Teradata customers are prime examples of this. We have clients who have had their Teradata systems for a dozen years or more only recently launch data governance for the first time. These data warehouses have driven tremendous business value, but as adop­tion has become more widespread, it’s time to formalize the policy-making around all of that data.

Conversely, with MDM, you’re harmonizing data for both operational and analytical purposes. Sometimes a data warehouse isn’t even in the picture. If you’re going to operationalize master data, you need to establish processes for sanctioning authoritative definitions and rules around that data. So data governance is really a mandate for MDM.

Why are BI teams uniquely qualified to start data governance?

They understand the complexity of integrating data. Sometimes the data warehouse is the only system in the entire company that integrates heterogeneous data from disparate systems. Data integration skills and processes are specialized. BI professionals fundamentally get the need to share common data, and they know the work it takes. After all, if you’re not sharing data, you really don’t need data governance.

There’s still a lot of talk about the need for executive spon­sors. What’s their role in data governance?

There are a lot of misconceptions about it, and often we find people expecting way too much. It’s sort of like a mythical deus ex machina, where a god appears from the heavens and fixes everything. In reality, the main job of the executive sponsor is to represent the overall busi­ness value of an initiative, and to then align the team’s efforts and goals around a common vision of success.

That sounds like a big job.

It is, and that’s not counting his or her day job. The best executive sponsors are in the throes of meeting their own professional objectives. They see how data governance can help them succeed in their roles in the short term, and drive enterprise value in the longer term.

So how is Baseline’s work changing as a result of the evolution of data governance?

A lot more companies have started on the data governance jour­ney than even a year ago. So in addition to our standard design and roadmapping services, we’ve just added a data governance maturity scorecard to our bag of tricks. It helps companies gauge themselves against best practices.

That service sounds like an idea whose time has come.

We could say the same about data governance!


Your Comment:
  
Your Rating:

Comments
Fuzzy Logix