Banks Can't Fix Data Problems After the Fact

The corporate suite has a new member, the chief data officer, who is tasked to oversee a relatively new function, data governance. As CDOs beat the drum about the quality of banks' data, a question comes up that accompanies any push for change from a new voice within an organization: Is anyone really paying attention?

These data quality overseers have the challenge of convincing their businesses that data quality will improve the bottom line. It is a significant mountain to climb. But the CDOs who will succeed and the banks that will make a real dent in improving data quality can look no further than the recent history of the automotive industry for how to proceed.

The decline in market share for American car companies starting in the 1970s – corresponding with the rise of Japanese automakers – has been well documented. Much of this shift in competitive advantage had to do with quality – of automobiles instead of data.

Since the time of Henry Ford, the production system of U.S. car companies barely changed until the 1990s. The focus had always been on volume: Get the product out the door, fix problems later. Stopping the production line for any reason became a cardinal sin. An engine put in backward? Fix it later. One factory, in Fremont, Calif., the subject of an episode of the radio program "This American Life," stood out in the 1970s and '80s for its quality problems.

Cars with defects littered the lot. Workers were paid overtime to fix them. Repairing a car after it had been assembled is an altogether more complicated proposition than fixing it when the mistake is first made. You need to take it apart and put it back together.

There were other problems granted, but the biggest issue for U.S. car company customers by the 1980s was reliability.

Enter the Japanese model of production. The Japanese system turned the Henry Ford production model on its head. Rather than focus on getting product out the door, whatever the quality, Japanese managers focused on quality first and then quantity. What did that mean in practice? Fix problems when they occur. Stop the production line if need be; stop that error in its tracks and make it right.

U.S. companies probably had seven times the market share of Japanese companies in the mid-1980s. The rest of course is history.

But what does this have to do with banks? Just like automotive companies, banks are making a product, whether it is a loan, a trade or a risk model. Based on the evidence of the daily financial news, from stress test failures to poorly constructed trading risk models, banks suffer from quality problems too. Given that backdrop, what can banks learn from the automotive industry quality issues?

First, it is clear that quality issues can hurt the bottom-line market share in a decisive way. Second, the case study illustrates the fact that the most effective way to improve quality is to attack the problem at its source. In the case of data, banks are currently expending much effort in assessing data quality right at the end of the transformation chain, whether it is in a regulatory report or a client statement. Fixing it at that point is a bit like fixing the car after it has already been put together. It is late and expensive to do.

Many hours of rework were avoided by the Japanese car companies, and can also be avoided by banks if they focus on entering data correctly in the first place. Assigning explicit responsibility for entering data correctly for new transactions and client relationships is key. Aligning performance assessment with the achievement of that task should follow. Third, improving quality in these ways can only be done by changing a bank's culture in a meaningful way. That may mean using case studies like this one to help people understand why data enhancement is important and align incentives accordingly.

Instituting this type of change is the name of the game for the first generation of CDOs. And banks that can get this concept will, like the Japanese car companies, be well-positioned to compete.

Andrew Waxman is a consultant in IBM Global Business Services' financial markets risk and compliance practice and can be reached at abwaxman@us.ibm.com or on Twitter @abwaxman. The views expressed here are his own.