"Club ties and long lunches [have] been swept aside" to make room for reformers, says the editorial. As proof, the magazine cites a Booz & Company survey that shows less than 12 percent of incoming CEOs were also made chairmen in 2009, compared to 48 percent in 2002.
That leaves a lot of entrenched chairmen/CEOs in place, but I'm not here to accuse anyone of impropriety. The point of interest is to acknowledge that one part of good governance is to segregate conflicted constituencies that would otherwise undermine the process. That would still sound smart had SarbOx never happened.
Like corporate governance, data governance works in proportion to how well we deliver what we profess to embrace. Data governance has earned a lot of serious attention and real commitment, but being relatively new turf for a lot of companies, it will also run up, silently or loudly, against conflicted constituencies. And yes, we will learn that different pockets of the business have vested interests that come at the expense of literal facts.
We can start with sales and operations managers who won't rush to see their performance lowered by standardized or otherwise reformed numbers. Like the banks singled out by The Economist as "odd creatures" that haven't been curbed by external monitoring, the performance compensated side of business is going to protect its interests, even as they are asked to set, own and manage parts of the data governance process.
Internal credit and risk departments -- just for starters -- are sure to interpret the meaning of governance differently. Figuring out how to have all these parties addressing policy, data definitions, ownership, rights and access at the table with a proper vote of influence is very tricky.
What is scary is the thought that some organizations actually will deliver viable and even comprehensive data governance policies that don't help performance in the end. The editorial points out that some banks that held up best during the financial crisis were no paragons of corporate governance, while some others with visibly outside directorship were among the biggest losers. It also mentions research that found no correlation of fiscal safety to governance except an inverse relationship between the number of independent directors and stock returns.
The authors of this study explained that this was because institutional owners had pushed for more risks that would lead to higher returns. I wouldn't bet that kind of lack of foresight or risk indifference would fare well against the mundane but effective checks and balances in brick-and-mortar world. But you can stop to identify where the interests of these constituencies lie and then consider what role they play in data governance. It's good reason to look after policies and metrics that neither allow aggressive goals to subvert governance nor unfairly undercut the performance of core functions of the business.
If you have heard of a framework or set of policies for data governance that considers the politics of performance rewards and risk -- or any observation -- feel free to share it here.
This is a very thoughtful piece on this widely talked about subject. Thanks.
As you pointed out, the purpose of any form of governance is to provide a framework for working out different vested interests. For a long time, our industry didn't recognize or confront the fact that when it comes to data, there are many powerful vested interests inside an organization. I'm glad to see this talked about in the swell of conversations on data governance.
I also agree with your point that there is a real risk of data governance not producing benefits. If it introduces additional cost, limits agility, then it could actually be very damaging.
When we developed the framework behind our software product, Data Governance Director, we declared some basic principles:
1. Data governance is about having the right policies for data.
2. The policies need to balance central control and local autonomy. Local could be a geography, line of business, or functional area.
2. Data governance is the BUSINESS PROCESS of defining, implementing, and enforcing these policies.
3. Data policies need to sit on the intersection between data and business process -- otherwise it'd be very hard to know whether your policies are truly beneficial.
4. Policies need to be measurable through metrics.
Our data governance framework is described here: http://blog.kalido.com/a-simple-framework-for-data-governance/.
Winston.
I also don't think that changing CEO's will affect corporate governance since the culture is what governs, unless the CEO is a dictator. Ken Lay, Jeffrey Skilling, Madoff and the numerous other corporate leaders who have been paraded before us along with statements like "we're doing god's work" are the reality in business. Regulators attempt to control the abuses but these controls are always after the fact. The damage has been done. Besides, these organizations spend a great deal of time in avoiding the regulations in the first place. It's a rather vicious circle.
Adopting the term data governance from corporate governance could result in the same consequences; data management avoidance. "We have a data governance committee and policies. They meet and discuss al there is to know about data. Therefore we are practicing data governance!" The term is like customer excellence and firms install IVR's to handle customer complaints (your call is important to us please wait).
Besides data governance implies bureaucracy (read government). Private organizations abhor bureaucracy so why use a term that conveys bureaucracy. Besides, I still don't understand how you "govern" data. People are governed. Another problem with the term. Is data governance a Trojan horse and an attempt to take over IT or other departments? Am I paranoid or is this data governance effort nothing more than an attempt to circumvent the power of IT?
Why not just call it data management. The term implies some degree of control, oversight and action. That is what we are really talking about. Deploying data management best practices.
The term data governance is pretentious at best and fictional at worst.