On Data Quality Sponsors

Sep 29, 2011 by Phil Simon in Data Quality, Master Data Management

In The Secret to Long-Term Data Quality Survival?, Dylan Jones writes:

Practically every book I’ve read on data quality states that you need a great sponsor on board. Some even go as far as saying that you are doomed to failure without one (although I’ve featured a number of people who have soldiered on regardless and had varying levels of success).

In my experience, both from working for companies and selling my own services, if there is no senior sponsor with enough “clout” to make things happen, you’ll fail to survive long-term.

I couldn’t agree more and, in this post, will chime in on the need for sponsors on data and system migration projects.

Agency Theory

Think about data quality from an agency theory perspective. In a nutshell, each party in an organization will tend to do what’s in its own best interests. That is, as a general rule, the marketing folks only care about relevant market share data. Payroll people just care about their numbers. Ditto finance and sales. While understandable, this type of provincial mentality often leads to significant data problems–not unlike some of the odd consequences of self-interest-based behavior described in books like Freakonomics.

Because we all tend to focus on only those issues that affect us, our paychecks, and our departments, the need for a data quality sponsor at large organization is difficult to understate. (Perhaps that person is the same as the Chief Data Officer, as Tom Redman mentioned on this site a few weeks ago.)

What if the Grand DQ Poobah:

Can broach MDM issues ignored by different lines of business in the midst of a time-crunched project?
Stops a data purge because of a potential downstream impact on other systems?
Routinely talks to other pockets of the organization about their data needs, suggestions, and struggles?

Isn’t this role worthy of at least significant consideration in your organization?

Simon Says

Regardless of moniker, in my experience, someone needs to maintain a global view for all of an organization’s data. What’s more, that person has to have teeth; making nice but ultimately futile recommendations does little good.

Feedback

What say you?

For more information about Phil Simon’s new book, click here.

Tags: agency theory, Data Quality, data quality sponsors

One Response to “On Data Quality Sponsors”

Dylan Jones

Sep 30, 2011

Great point mate, I often think this role should be titled “The Overlord of Data Common Sense” because most decisions that are taken in big organisations defy any logic.

One example sums this up…

Very big utilities company developed 15 systems over a 10 year span to hold broadly the same type of equipment. They developed 5 different standards for naming how these same pieces of equipment should be identified at a particular location.

And after 10 years they realised that this was a bad idea and decided to launch effectively an MDM mega hub to bring it all back under one roof.

To me, a Chief Data Officer should have oversight on this to say, hang on, 15 systems, 15 lots of developers, 15 lots of DBA’s, 15 maintenance contracts, 15 servers, 15 sets of users, 15 office locations, 15 sets of training guides, 15 sets of nightmare feeds and horrendous data quality issues to reconcile, 15 sets of replicated master data.

Is there a better way?

The CIO obviously signed off all these projects but these guys often just see functions, not data, “yes we need these 15 functions, I’ll sign off”.

You need a CDO to predict these ridiculous events taking place and actively roll back the nightmares that have gone before.

Great post, like those examples and I feel a follow-on, follow-on post in the making!

Reply to this comment