Rants
Questions
Soapbox
Best Practices
Apply today for a FREE subscription to CIO Magazine!
Wed, Oct 28, 2009 5:26 EDT
|
Posted by: KalidoCEO in Best Practices Topic: Enterprise Management
Current Rating: |
In a recent column, Dan Brown talks at length about the issue of data ownership. He contends that most C-level executives don’t know if their data is accurate. The fact is, most business users could look at data and think it is “accurate” but not understand how it can be “bad” data. What is really required is for business users to be intimately involved in the definition of their data, the business processes that use it and the applications that consume it.
Much has been made of data quality, but much less of data context. Most companies focus on their largest data sets, like customer or product data. But how much is really known about the data that is used to make decisions? While it certainly doesn’t help to get a customer transaction wrong, the actual economic impact on the business can be small. What if your pricing data was corrupted? The domino effect could be disastrous to your business, not only in the short term but the long term as well. The fact is that unless the business is directly involved in determining what data is to be generated and its value to the business.
This is often lumped under the category of Master Data Management, but it’s so much more. It’s the definition of rules and policies that govern data and how it is utilized throughout the enterprise. It encompasses transaction data, reference data, AND master data- it requires a keen understanding of 3rd party data used as well as data that is created and maintained outside of a transaction system. If you truly understood the value of the data you were working with, would you treat it differently?
Let’s look at this example; a product brand manager has to determine the promotion schedule for their product for the next 3 months. The PM will look at sales over the past 3 quarters to spot trends; changes in demographics and buying patterns; the price elasticity of their product compared to competitive or replacement products; the forecast from sales; the report from production outlining an increase in raw material costs; their personal compensation targets; and most likely how other brand managers in their company have been doing.
So, where does the PM find this data? By my count, in about 50 different places, from spreadsheets to transaction systems to external data to some they got from the guy in IT they play pickup hoops with. The PM may or may not get the answer they are looking for, but data only satisfies one dimension. The worse the data, the more reliance on gut instinct, the better the data, the more accurate the prediction.
The other factor here is who else uses the data. The same information used by the PM may be just a slice of what is needed for production, sales, finance, etc. However, because so much of this type of data doesn’t exist in a managed system, it is unavailable to the other departments.
Business users need to see how their data affects the company as a whole and play their part in delivering high quality data that can be used enterprise-wide. In order to understand it, they need to be made a part of the team that defines, monitors and manages the environment. Companies doing this today are seeing breathtaking results in efficiency and effectiveness. This concept of data governance is now a reality that companies must embrace starting today.
Excellent post. I completely agree with the sentiments expressed.
When asked what does ‘Data Quality’ mean, C-level executives typically respond along the lines of ‘the data is either good (accurate) or bad (inaccurate)’. They have little understanding of the commonly used dimensions of data quality.
The above is one of 10 common enterprise wide data Governance issues I have identified on my blog (far from an exhaustive list). I also provide a process you may use to assess the status of Data Governance issues in your Enterprise or that of a client.
I hope you find it useful
Ken