Measuring Data Assets — What is the true cost of bad data?


You have to know the business value of all your data assets to get a good return on your investment.

Bill Hewitt

Recently, a business executive told me that they expect 20% of their customer data to be “bad” at any point in time, a cost of doing business, according to him. Like many business theories, most managers have come to accept certain myths as truth; that you can’t have growth and profitability, that employees are generally lazy, that competitors will act ethically to form a healthy market. Not only is bad data bad for business, it can be catastrophic.

For years, data quality vendors have talked about the impact of bad data, usually classified as incorrect data. The classic example of direct mail being returned is used over and over. What doesn’t come out in this example is the true cost of these errors. Sure, it would be great to avoid the cost of mailing, but what about the opportunity cost of not reaching new customers? If your direct mail cost 50 cents, and you have 5,000 pieces returned, your “cost” is $2,500. However if your average new order is $45 and you convert 20% of your mailings, then the opportunity cost is $45,000.

Now imagine something far more unsettling. You are a pharmaceutical company and need to inform your customers that a certain drug has violent interactions with a new popular drug on the market. If 80% of the data you hold on 100,000 physicians is good, that means 20,000 doctors will not get your warning in a timely manner. They could be sued by patients, you could be sued by the doctors and ultimately, people could die.

Do companies really understand the value of their data? Most companies equate value with volume, but real value is determined by the impact on revenue, cost and risk of that data. This is why I believe data should be valued as an asset; some data is simply more valuable. In some cases, data can be used to create a significant competitive advantage.

So how do you know which data is good and which is bad? It all starts with the business process. Think about a simple one, like order to cash. What’s worse- sending an invoice to an incorrect address or sending the wrong invoice amount? If I send the wrong invoice amount, say a lower amount, chances are the customer will pay it and I’ll never know the difference until an audit compares the invoice to the contract. If I send to the wrong address, I’ll get the invoice returned. The cost of having an incorrect invoice amount can impact not only my profitability, but also my customer satisfaction.

The good news is that most companies know what their business processes cost, so it’s easy to figure out the tangible economic impact, but the risk to the business or lost revenue is more difficult. To figure this out, use a similar scenario and build assumptions as to what happens after impact; will my customer go somewhere else? Will I face regulatory fines? Will my company get distracted by us being sued?

The case for managing data more effectively is real. Data Governance is the process that gives you the capability to define, manage and monitor data policy across the enterprise. It serves as a compliance system, an early warning system and ultimately, an economic measure of the financial value of your data asset.

Tags: , , ,

One Response to “Measuring Data Assets — What is the true cost of bad data?”

  1. residential cleaning prices June 20, 2010 at 2:43 am #

    Since there are many blogs and forums with unique points of view, they test your comprehension. It really is at these moments when you have helpful insignt other people might not have had, not to mention the blogger her/himself.

Leave a Reply