Problem Solved. Now What?
We’ve solved enough problems; let’s implement the solutions at hand.

A. Blanton Godfrey

We in quality management may have been working on the wrong problem for years. We’ve focused on creating and teaching better statistical methods. With advanced software and popular methodologies such as Six Sigma, we now teach these programs to millions of people in all kinds of organizations. But most organizations still deal with incredible numbers of problems; simple problems remain unsolved, and many see the same problems repeatedly popping up.

Some years ago, a group of my quality management graduate students at Columbia University worked on the distribution problem for local delivery of The New York Times. The team collected sales data on the Sunday Times from newsstands during the year. Data analysis showed that sales were seasonal and predictable. By calculating some simple distributions using the time of day when the newsstands ran out of papers on both high and low sales days, the students predicted how many papers could be sold on these days to eliminate wasted copies. The evidence was strong, but as far as I know, no one ever used the results.

Another group analyzed checkout times at a local grocery store. The root cause for long lines became obvious when the data were analyzed. Clerks new to the cash registers took far longer than those with more experience. Throughput could actually improve using fewer registers and clerks. However, the store’s high turnover of clerks meant that most registers were staffed with inexperienced employees. The reason for the high turnover was also easy to discover--low wages. The math was simple: Increase wages, lower turnover, reduce staff and increase profits. It seemed so obvious. But the store manager was unimpressed by the numbers. His supervisors were more concerned with keeping salaries low, so he was, too; improving profitability by raising wages wasn’t important to them.

Time after time in consulting projects, I’ve seen the same thing happen. Analysis leads to a clear solution, but the solution is never implemented. Something in the company culture prevents people from facing, or perhaps using, the facts. Instead, we continue doing things the way we always have.

Some years ago, I worked with an organization with 21 service sites. It desperately needed to improve throughput because it had far more demand than capacity. The organization experimented with one pilot site and increased the number of customers served per day by more than 20 percent. Problems arose, however, when it tried to replicate these results across the other 20 sites. Managers presented numerous reasons why the same changes wouldn’t work at their sites. They questioned the data and the manner in which the study was done. Ultimately, the company never achieved the same results at any of its other sites.

A similar stalemate occurred in a different service company, this one with 250 sites across North America. One improvement team managed to reduce by nearly 80 percent a major problem common across all sites, and in doing so it saved more than $300,000 at that site alone. The CEO, seeing a potential $75 million jump in profits companywide, demanded that the other sites implement the improvement as well. Nothing much happened. Some sites made some progress, but most found many reasons why the solution wouldn’t work for them.

As quality professionals, we’ve been teaching problem solving for years, but perhaps we should have spent more time teaching solution implementation. Too often we take the easy way out and blame management for failing to implement profitable solutions. It’s easy to cite 10 or 20 reasons these ideas won’t work in our organizations, but it’s much harder to discover 10 or 20 ways to make them work.

In both of these organizations, the improvement projects were carried out almost secretly. Management assumed that if positive results were achieved at one site, all the others would want them as well. This wasn’t true, of course. Site managers didn’t believe the results. Their measurements were different, and they couldn’t translate the improvements to their sites. Senior leaders were supportive but couldn’t address these objections. They didn’t understand that each site manager would also have to understand the methods, tools and processes by which the results could be achieved. People from each site needed to learn from the pilot sites and modify the approach to fit their sites. They needed the same support the pilot site managers had provided their teams.

The real key to implementing improve-ments across an organization isn’t merely leadership but also widespread active involvement. This requires knowing when to provide the right support at the right time. In both the examples, the organizations should have created steering teams with representatives from every site. They should have agreed on the improvement plan, which measurements to make and how they’d be modified to fit all sites. The improvement processes across the sites should have been standardized to make comparisons easy. Everyone involved would have to know that differences would crop up from one site to the next. But each site would have been able to achieve similar results.

About the author

A. Blanton Godfrey is dean and Joseph D. Moore Professor of North Carolina State University’s College of Textiles. He’s co-editor of Juran’s Quality Handbook, fifth edition (McGraw-Hill, 2003), and co-author of Modern Methods for Quality Control and Improvement, second edition (John Wiley & Sons, 2001).