Guest column
How to spring clean your organization’s data
By Connie L. Van Fleet
Senior Vice President for Information Services
Chief Information Officer
Blue Cross Blue Shield of Missouri
It’s time for a thorough spring-cleaning of your company’s data. Corporate data are a lot like someone’s closet: packed to the point of bursting; filled with items that are similar but not quite identical; archiving things that haven’t been worn in years; sprinkled with a few items that were trendy but aren’t useful over the long term; holding other gems that are worn all the time because they’re good quality, they work well together, and they fit your lifestyle; and storing the indispensable favorite pair of bluejeans.
Look around your work area and your company. There probably are stacks of paper or piles of reports sitting there. Some of them may be yellowed and dusty. Or if you’re the "paperless" company, ask your CIO how much data he’s archiving that never get recalled. Think about what that costs.
If your organization is like most companies, you’re drowning in data and starved for information. Just like my closet; stuffed to the max and yet I have nothing to wear. The road between a sea of data and high-value information has three rest stops: vision, a pragmatic approach, and a commitment to quality investments. Oh yes, there’s a lot of hard work.
Rest stop one: Vision
Vision is the magnetic north. If I’m cleaning out my closet, or cleaning out my company’s array of data, I’m only going to be successful if I have a vision of what I want the final product to look like. If I don’t, I’m just going to keep repeating my prior experience. And here’s where many companies really struggle: They can’t achieve a common vision. People come and go. People are reactive to the current crisis. People solve only part of the problem and then forget the rest of the work. But vision implies a view of things as they should be in the most perfect world. Vision also implies something that you keep rechecking, just as we orient ourselves from that northern indicator on the compass when we’re lost.
My own company embarked on a vision for analytical reporting a number of years ago. The principles of that vision included building a single repository of data, integrating all the source transaction systems into a common view, and delivering the data through self-service distributed reporting. The vision was to get the data close to business users, and then they could create meaningful information.
Rest stop two: A pragmatic approach
Taking inventory and knowing what you have is the starting point. Doing the inventory is not so hard, but knowing what the data really are might be. Ever had a blouse that looked white in some light, but ivory in others? Or a pair of pants or socks that could be dark navy, but also might be black? You get a rude awakening in the middle of the day when you realize that you have one black sock and one navy sock on. But that same experience happens frequently in business. That report you thought contained the complete medical expense for a product line isn’t complete. Why? Because you might have said you wanted the claims expense for the product line, and the analyst interpreted that literally, excluding capitation costs or fees paid for care management programs. You have data. You don’t have useful information because you didn’t get what you thought you asked for.
Ever had the experience of attending a meeting, and two departments have analyzed the same business question and provide two completely different conclusions? They’re both coming at the task with their own understanding of the data, and most likely their own separate sources.
So taking inventory and knowing exactly what you have is crucial. If you understand the vision and you understand what you have, then you have a chance to architect your data into useful business information. But what’s missing?
Rest stop three: Quality
Quality is the anchor of your vision. It’s the bedrock, and if you don’t have it, you will eventually drift into danger. You’ll be back out in the sea of data.
I mentioned that my own company started with a vision for an enterprise-architected data warehouse as the solution to business analytic and ad hoc reporting. We built it, they came once or twice, but didn’t come again. Why? Because the quality was inconsistent. Business users got burned basing critical business decisions on the data because we couldn’t guarantee the quality.
So we invested in a process for measuring and ensuring a level of quality in our analytical data. We developed the key data element (KDE) measurement process. First, we partnered with business units to identify the critical data elements that drive the majority of business decisions for pricing, product design, reimbursement methods, and other corporate expenses. Then we developed a set of statistical measures to evaluate every month’s data load against a set of predefined minimum tolerance levels to determine whether the data are up to standards. If not, teams work on correcting the problems at the source systems if possible. This process has provided better end-to-end cooperation between those teams accountable for transaction processing and those teams accountable for analytical reporting. And it puts a consistent product at the fingertips of my business partners.
Sometimes in your quest for quality information, you need to revisit all the rest stops. Recently, my company modified its vision to add another element. We found we’d been fairly successful in the distributed reporting aspects, especially since we improved the quality aspects. But that still left various departments either information-rich or information-starved. It overlooked the needs of some of the senior executives. It also caused duplication of efforts among different departments with slightly different outcomes for similar efforts. Remember all those things in your closet that are fairly similar, but not identical?
We’ve modified our vision to include developing a standard set of metrics and reporting capabilities focused around key levers that drive the pricing of our products and targeted to senior executives.
Yes, there’s the natural desire to ask for everything during this project. But we’re winnowing things down by asking the question "So what?" — meaning what are the actions we’d be able to take if we had that information. If it doesn’t result in an actionable step related to our key business drivers, we drop it from the project scope. We’d just be filling up our closet again with useless stuff.
If you’re like me, every time I clean out my closet, there are some things that always stay. Usually, those are the clothes that I really thought about before I bought them. They’re the ones that were high-quality and not just trendy. They’re the ones that match most closely to my lifestyle needs. And they’re the ones that provide the most value to me year after year.
Your company’s closet of data is not much different. It needs a good spring-cleaning, too. Look for the quality stuff, and don’t forget to keep the bluejeans too. Everything else can go.
[Connie Van Fleet can be reached at Blue Cross and Blue Shield of Missouri, 1831 Chestnut, St. Louis, MO 63103. Telephone: (314) 923-6273.]
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content