In case you haven’t seen the nice data decay infographic from ZoomInfo, I’m reposting a smaller version here.
Essentially, the graphic is showing the level of change of master data—and only for business information. Understanding data decay is essential, because many times organizations think that if they establish a huge, one-time effort to clean up their key master data elements, they won’t need to do it again.
Not true, as the data decay infographic shows. Your data can decay over time. Addresses and zip codes are always changing because of postal authorities. People move and change jobs frequently, as described above.
Mergers + acquisitions
One key cause of data decay is mergers or acquisitions. Are there any mergers or acquisitions happening in your industry? :-) Check out the situation with Sara Lee.
Sara Lee renamed its meat business to Hillshire Brands, while the beverage business will be D.E Master Blenders 1753 (which will also be merged with CoffeeCo).
Now let’s think about the kinds of activities you are doing with this type of business master data:
- Optimizing supplier channels
- Optimizing logistics costs for suppliers
- Selling product to the vendor
- Marketing new products to the vendor
- Measuring customer satisfaction
To do any of the above tasks, you need a time-based understanding of the company “Sara Lee”, when pieces of its brand were renamed, when pieces of the brand were split, etc. Otherwise, how will you be able to tell if you are doing more or less business with “Sara Lee” this year compared to last year?
The answer, of course, is that you need to establish an organization which understands that information is an asset that can—if bad—be a silent virus. The organization may think they are making good decisions and transacting efficiently, but decayed data is silently making these tasks much less successful.
You’ll also need to establish core quality metrics of you can apply to high priority master data elements. Operationalize these metrics so they run on a recurring schedule. When the quality dips below a certain threshold, you can initiate a corrective program *in advance* of a raging data fire.
You’ll also need a data governance organization that can manage the uses of time-based information so you can inform business analytics.
Are you there? If you are not sure, try checking out the Information Governance Benchmarking and Best Practices survey.