Data Governance

The last living man born in the 1800s died earlier this week. Data governance programs worldwide can now change their business rules, losing hope that Misao Okawa might become a customer. From this point forward, any new Person clients that are Male must have been born after 12 September 1901 (just in case James McCoubrey opens an account).

At a minimum, the DG “message” should consist of the following: What is DG? Why do we need it here? What are the benefits of DG? (Try to tailor this to your audience)

Tina McCoppin addresses success factors for a data governance communications plan. Marketing data governance is critical and Tina highlights some core messaging opportunities. Her minimum list is the brand (the what), the motivation (the why), and the benefits.

The motivation and the benefits are two angles on the same fundamental message. As I was reading her advice on messaging, it reminded me of Simon Sinek's Golden Circle, which captures how creative innovators focus on the "why", the "how", and then the "what" in that order. They focus first on their motivation and let that drive the "how". The "what" becomes incidental. This applies to data governance as much as to an iPad. One of my current priorities is to tune my ability to quickly and substantially articulate the "why" of data governance for different audiences. Data governance can improve customer engagement, reduce operating costs, increase revenue, and even increase the value of the company. It can reduce employee frustration as a fringe benefit of improved data quality. It's important that data governance leaders articulate these benefits, tailored to the varied audiences of investors, executives, data owners, and data stewards.

Tina rounds it out with a robust laundry list of important messaging points related to:

data governance policy

metrics

business case

status

stories

media coverage

One of the biggest challenges I think we face today in Data Governance is the big bang theory – the idea that we can get everything from nothing very quickly.

Janine Joseph write about how establishing data governance is more of a gradual journey than a Big Bang. She closes with these observations:

Data stewards are not optional

Metadata matters

Data quality can't do it all (echoing a comment Genia Neushloss made to me earlier today)

Master Data Management Will Succeed With Stewards, Metadata, and Data Quality

Job/Role Descriptions Are Required

I concur with her points and my recent activities launching a data governance program mirror these points. Things I did today include:

Finalizing a RASCI for our data governance roles

Met with a leading metadata management platform provider

Reviewed options for business rule statement standards

Made an appointment to follow up with an MDM platform provider

Early in an organization’s data governance maturity, many practices require some assembly, and since necessity is the mother of (not only good) invention, not-so-good inventions often get adopted as best practices in part because of the excessive admiration of that self-assembly.

Change agents get flummoxed when proven industry best practices are resisted. The confused looks on their faces seem to say: “Can’t these people see how wobbly their bookshelves are?”

Jim Harris discusses how we tend to have deeper affection and admiration for what we build ourselves, including data governance processes and capabilities early in the maturity cycle. For those of us building new data governance programs, this gives us two questions to ask ourselves:

1 - Are we aware of the existing IKEA effect and the emotional entanglements with existing data governance processes and artifacts, e.g., that business glossary on the intranet that represents the blood, sweat, and tears of data governance pioneers that needs to be moved into a formal business glossary and MDM solution?

Gartner Group's Svetlana Sicular discusses the convergence of big data and data governance and her experience that many big data conversations gravitate to data governance.

Bill Meinweiser, Senior Delivery Lead, Data Governance- Americas for Utopia gives his overview of data governance, its structure and best ways to tackle the beast.

From the 2013 SAPPHIRE NOW + ASUG Annual Conference, the Utipoia team addresses four data governance questions:

1.  Data governance is thought of as both a technology and a process. Can you differentiate them?

2. What does a data governance organization look like?

3. Talk about the role of the process in data governance.

4. It seems the most common question about data governance has to do with tackling everything at once. Is that the best approach?

Before you put a big data business case together, spend some time trying to understand the data and its content, in order to comprehend the ROI. This is where the data governance aspect comes into play

Laura Reeves discusses data governance of data outside relational database systems, such as content management systems (CMS). She notes that tagging and classification processes performed manually in the past will not scale in the emerging world of Big Data, requiring more automation.

With the promise of big data (solving the unsolvable problems, informing better decision making, creating new products and services, discovering patterns and acting on them, etc.) on the horizon, what has really changed?  Does this mean that everything we know and do with not-so-big data should be tossed?

This is why data governance needs Searchers, not Planners.

http://ciowhitepapers.com/owp/169

Co-authored by Joyce-Norris Montanari (@JMontanaria - a former colleague) and Manish Sharma, this is one of the most useful 12-pages I've read on data governance. Starting with the obligatory, but useful, definition of data governance and a comparison of data governance to data stewardship, the paper moves on to a robust data governance maturity model. Best of all, the authors provide concrete steps for moving from one maturity level to the next.

If data isn’t managed well, your organization could be at significant risk. This list will help you spot areas where your data governance may fa

Some of these problems look like success if you just look at the major points.

1: You have pockets of adoption

You always start with pockets of adoption, even when you have an enterprise-wide initiative. Don’t try to boil the ocean.

4: There are multiple data stewards

There are nearly ALWAYS multiple data stewards when there are conflicting business rules about the data. Identifying multiple data stewards for the same data is the first REQUIRED step to resolve conflicting business rules.

Data quality is where it’s at