Click to learn more about author David Schlesinger.
Many people have written about the importance of Data Quality and most everything they have said is quite true, but besides the obvious costs of poor quality data. There are subtle hidden costs as well.
A few of the obvious costs are the need to manually configure downloads of data from source systems for every single new software project, forever. This is often due to lack of a trusted and complete enterprise model with detailed business metadata – this additional work to redefine the data is part of every project and therefore done again, again, and again.
Another less obvious cost comes from sales that do not happen. When the inventory displayed to a customer on a webpage does not show all the potential products that are for sale, the items not shown are not sold. This lack of data may be simple as a gap on an order page, or as irritating as a customer web order that informs the buyer only when at checkout that there are no longer any blue jackets size XXL left in the warehouse. The item appeared on the web page as available, but at purchase time was actually not available. (Worse, it may have been actually available but the inventory system did not update the web pages.)
Not only did the sale not happen, but also the customer longer trusts your webpages to supply products. He or she will often start buying at a more trustworthy source elsewhere. When this happens, you receive no notification. It is just that you continually have fewer repeat customers.
This is the key problem with subtle losses; lost customers seldom tell you they now buy from a competitor. They just go away. Surveys and such may or may not capture this because folks seldom complete surveys for companies they no longer use. They may not even remember why they left, or why they no longer trust your company. But sales from them disappear due to a Data Quality problem.
Another subtle loss occurs when managers find their reports are suspect. If one business reports that sales are at one level, another group gives different numbers, managers may choose to wait and view a series of reports over time to make sure that each supposed market trend is real. Untrustworthy reports introduce decision latency.
This is an opportunity cost squandered by arriving late to every new marketplace trend and change. It is an advantage given to all their nimble competitors. Why are some companies nimble? Perhaps because nimble companies trust their report data. Trusting your data requires trustworthy data. Trustworthy integrated data is a competitive advantage.
Another example of subtle business losses due to poor quality data is where companies keep separate customer databases by department. Addresses change all the time, and a person will make an address change only once to an organization. After that, the company is expected to know where they have moved. This is true for corporate customers, as well as individual retail customers.
One way to find if you have this cost is to investigate mailing and shipment returns, then then tabulate the employee time it takes to rectify a poor address and the cost of re-shipping the product. (Finance may not count employee time as an extra cost, but that is mostly reflection of how they do tax accounting.) If a company has 100 employees, but half of each day they are doing what Larry English accurately called “scrap and rework,” They are wasting money and resources.
The cost of poor customer data is increased shipping costs, a waste of employee time, and late customer delivery. When you are spending more money to degrade your customers experience it is a lose – lose for all.
These and other reasons have been mentioned elsewhere, but they remain solid business reasons to improve Data Quality because it is directly attached to the company’s bottom line as well as the customer’s heart.