Continue in 2 seconds

DQ Point 11: Eliminate Numerical Quotas

Published
  • January 01 1999, 1:00am EST
More in

This is the sixteenth in a series of discussions of quality guru W. Edwards Deming's fourteen points of quality and their ramifications for information quality. Here I describe Deming's quality point 11, "Eliminate Numerical Quotas," a requirement for enabling sustainable data quality improvement. A senior vice president of a health care company hired a telemarketing firm to conduct a customer satisfaction survey of their top 50 customers. The customer list came from the customer service system and contained customer name, address, phone number and contact name of their most important customers. The telemarketing company called back to say they were going to have to charge more money than originally planned. "Your data is so bad that we are spending much more time on the phone conducting this survey than we planned," was the reply. "We are having to call directory assistance and then find the correct party within the account because your contacts and phone numbers are out of date."

The senior VP ­ who had not been particularly supportive of the company's information quality effort ­ sought to get to the root of this information quality problem. Asking the customer service center questions (such as, "Why don't you verify address, phone and contact information when a customer calls in?"), he found the reality of Deming's Point 11 of quality. He discovered that the customer service procedure, in fact, does say to verify the customer's address and phone number with every contact. But the reps were not following the procedure because they were measured by the number of calls they took per day. If this was the level of information quality for these "best" customers, what information quality can they expect of the rest?

Quotas Decrease Quality

"Quotas or other work standards...impede quality perhaps more than any other single working condition."1 The solution is simple. If management believes quality is important to business success, they will implement Quality Point 11: Eliminate numerical quotas for the work force and numerical goals for people in management.2 In their place, management will put measurements for such things as customer satisfaction.

The impact of quotas is counterproductive. Peer pressure causes those who perform "above average," whatever that means, to slow down. In essence, they produce less than they are capable of. Those who perform below the average cannot make the rate. The result is loss due to working faster than optimal, frustration and high turnover. The actual effect is to "double the cost of the operation and to stifle pride of workmanship."3

Furthermore, point 11 strikes against the practice of management by numbers. Internal goals, (such as "decrease costs by 10 percent next year" or "increase sales by 10 percent") "are a burlesque," Deming says.4 Meeting a goal may be due to a natural fluctuation rather than to a process improvement. Quotas are even more cruel if they are arbitrary and not accompanied with a plan for attainment. The only numbers that Deming believes are permissible are those that set forth actual facts of enterprise survival, such as, "unless our sales improve 10 percent next year, we shall be out of business."5

Ramifications

The information quality problem described in an earlier article in which an insurance company discovered that 80 percent of their claims had a medical diagnosis code of "broken leg"6 confirms the application of point 11 for information quality. The claims processors were measured by how many claims they process in a day. How can you expect them to take time to enter the correct medical diagnosis code when they could use the system default of "broken leg?" It should come as no surprise that 80 percent of the claims "just happened" to have a medical diagnosis code of "broken leg." While this "quality" satisfied the operational process to pay the claim, it was non-quality data for all other information customers, such the actuaries who needed to analyze risk from the data warehouse.

When quotas and speed cause information quality problems, it significantly increases the costs of the affected information products. The data is not usable by other business areas, so they have to look for alternative sources. Or they have to spend a significant amount of time and money to verify, find and correct the data before they can use it. Or they may even have to create and maintain their own private databases and systems in order to trust the data they require. All of these costly workarounds can be traced to the root cause that the information producers allowed non-quality data because they were being measured by how fast they did the work, not whether it was reusable by others who needed it.

The only viable correction to these problems is to change the performance measures for information producers. Rather than measures of "how fast" or "how many," use measures of customer satisfaction by the downstream information customers. Data captured with quality at the source will meet all knowledge workers' needs. This results in decreased costs and increased productivity of those who use the data. The data warehouse team does not have to embark on a major cleanup effort to find the right medical diagnosis code for 80 percent of the claims. The customer service database will be current, reducing the costs of conducting a customer satisfaction survey. And this was only one instance of use of the customer service customer information. Consider how many other processes failed trying to use this information. Or worse, how many private customer databases were spawned because knowledge workers who needed vital customer contact information could not trust it from the official and supposedly authoritative source?

Consider the economics of quotas that have the effect of decreasing information quality. Because others cannot use the poor quality data in a common sharable database or in managed replicated databases, they will be forced to:

  • Not use the data because it lacks quality. The result is sub-optimization of the process or decision due to "missing" (actually unusable) information. This is the equivalent of manufacturing "scrap." Lacking the necessary quality, and too costly to rework, it is "thrown away."
  • Look for an alternative source (supplier) of the information. Creating their own private databases, possibly by extracting the data from the original source, they clean it up for their use. This is the equivalent of manufacturing "rework." The further tragedy of this "solution" is that the newly cleaned data is not available to other knowledge workers because it is being maintained in a private, proprietary and inaccessible database.

One newly appointed information quality staff person confided to me that she had been "guilty" of developing her own proprietary database when she was in the business area. She was not able to use the data from the official corporate database because it lacked quality. The financial company's business drivers happened to be speed of work. Never mind the quality or the costs. The company must capitalize on new business opportunities quickly due to the competition. A high degree of redundant databases exist in this company because speed of product delivery to the marketplace superceded the "luxury" of quality, shared databases. But at what cost? Even as I write these words, the company is under pressure to reduce its costs of operations and has postponed some vital projects that could mitigate the costs of the redundant databases and un-integrated applications.
When information producers are forced to make a choice between information quality that others can use and quotas for which they are measured to receive their pay raises, the choice is already made. Only management can change the performance measures. When management recognizes it cannot afford the high costs of poor information quality, it will change the performance measures.

What do you think? Send your comments to Larry.English@infoimpact.com or through his Web site at www.infoimpact.com.

1 Walton. The Deming Management Method. P. 78.

2 Deming. Out of the Crisis. P. 70 & 75.

3 Deming. Out of the Crisis. P. 71.

4 Deming. Out of the Crisis. P. 75.

5 Deming. Out of the Crisis. P. 76.

6 English. "Data Quality: Meeting Customer Needs," DM Review. November, 1996.

Register or login for access to this item and much more

All Information Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access