Jill Dyche’s Talk at Gartner MDM 2013
I attended the Gartner 2013 MDM Summit, and had a chance to hear Jill Dyche, VP of Best Practices at SAS.
Due to recent trends, executives are finally getting serious about treating data as a corporate asset. In the financial services and healthcare industry, there are new trends and regulations. In the insurance industry, there are widespread moves towards IT modernization.
But executives are busy – how do you get their attention? How do you get them to understand the role that data plays in running the organization better. People still tend to think it’s an IT issue. How executives lead informs their perspective on IT and its business value.
Some executives see anything relating to data governance as a bureaucratic exercise. Jill referred to the “kick-off and cold cuts” approach. Everyone sits around the table, eats lunch and complains about the data. Six weeks later, there’s a meeting with no free lunch, fewer people show up, and it again turns into a complaint session. It turns into a people-driven exercise, at the expense of a process-driven approach that is focused on the need, pain or problem that we’re trying to solve.
As far as Return on Investment (ROI), Jill agrees that we can monetize data. But we often don’t have the “before” measures in place to measure the “before and after” ROI picture.
With the time and money that simple efforts like inventorying our data, what could we have done?
Part of the problem with data governance efforts that have failed, part of the problem is that data governance is a “squishy” term, with a lot of variation in terms and no unified vocabulary. We need to apply leadership to fix this, because it’s a cultural issue.
Jill had a great chart showing the evolution of data awareness, from the initial value proposition of data warehouses as “mainframe data offload” to CRM, to data quality and data integration, to data stewardship and organizational efforts at data governance, to process and policy driven efforts at data governance.
Because a lot of knowledge about our data is tied up in “tribal knowledge”, we need to move more towards a policy and process driven approach, so we’re not so dependent on a few people with in-depth knowledge of the business and how it uses data.
The first few stages in the evolution of data awareness are marked by a focus on Platform. The next few are focused on Integration. Only the last few stages in the evolution are process and policy oriented.
It’s all about selling the idea (and how hard this is) to the C-level executives. But it’s an uphill challenge, because the executives sometimes don’t have the attention span.
When you turn the conversation to “where is the business going?”, then it’s harder for executives to say “No”.
Jill gave an example of how a CEO at an insurance company defined their three corporate strategic goals. But it’s one thing to define corporate goals and objectives, but it’s another to execute on them.
If we take corporate objectives and distill them down into initiatives and strategies, we have a better chance of achieving them.
The map is from the corporate vision, to the overall strategies, to the key focus areas required to achieve those strategic objectives, which give rise to the initiatives required in each of those focus areas.
Then you look at the initiatives and see where you will require better master data to carry them out. So you wind up creating a strategy map for data that corresponds to the overall corporate strategy map.
This makes it very hard to say “no”, because if you agree that the corporate objectives are important and that certain initiatives require better access to data, then you have to agree that strategic initiatives around data are necessary.
But a data governance framework can really highlight capability gaps – things the organization doesn’t currently know how to do.
The question is “can we do data management without data governance?”. The answer is yes, we’re already doing a lot of those things (although it’s not ideal).
But you cannot do data governance without data management (if you did, it would turn into an academic exercise very quickly).
So the question is – if you had a good strategic roadmap and data governance framework, would you be able to execute?
The data stewards are the ones in the middle that are the interface between the business and IT.
Jill gave a good example of Harrah’s from the gaming industry. They were lucky in that they had the metrics from the “before” time period, and from after improving their data capabilities.
In terms of valuing data, you need to look at that in the context of its usage.
So where do we go from here? Data governance is not a series of meetings – interest erodes and people lose interest. There’s a trend towards “regimes” – where only a part of the group has to work on a particular problem. You can use the Responsible / Accountable / Consulted / Informed (RACI) approach to figure out who needs to be involved in any particular issue.
This brings people to the table. They realize that they’ll have a say in the data that is meaningful to them, without having to be involved in everybody else’s issues.
The “aha” moment is when you’re able to tie improving information management capabilities directly to corporate strategy. It makes it very hard for executives to say “no”.
Also, business intelligence is becoming a commodity. So BI is becoming easier and cheaper than its ever been. But data is getting harder and harder. Not only are we getting more and more data, but it’s coming from more and more places. Five exabytes were generated by business in 2008. Now, business are generating that amount every two days. Data is getting harder, while reporting and analytics getting easier.
Jill moved on to talking about building a comprehensive, sustained data strategy. Understand what’s important to management now. Tie it to your corporate goals, make it an ongoing program (not a project), look at where the company is investing externally – that’s a good clue about what the strategy is.
Work within your culture to avoid saboteurs. Some companies have the “culture of no”, where people aren’t rewarded for saying “yes” to new things. Data governance looks different at organizations that are top down vs. those that are bottom up.
Understand your current state before making the pitch for data governance.
Choose sponsors based on who owns the business problem. Focus on the need, pain or problem that we’re trying to solve. Don’t automatically go to the “friends of the data warehouse”.
Tie the operational metrics for data to the context of the business initiative.
In closing, let’s talk about criteria you can use to see whether your company treats data as a corporate asset:
- the asset has value
- the value is quantifiable
- the asset helps the company meet its strategic objectives
- it requires specialized skills to build and maintain
At a lot of companies, one or more of those criteria are in place. But the tough questions are: are you giving it resources comparable to your other corporate assets? Are you dedicating technology comparable to your other corporate assets? Are you allocating funding to data relative to other assets (does data have its own budget)? Are you measuring the cost of poor, missing or inaccurate data? Do you understand the opportunity cost of not delivering timely and relevant data to the business?
If not now, when? When we’re managing exabytes instead of petabytes?
Somebody’s got to do something, and it’s sad that it has to be us. But we’re the ones that know how to solve these problems.