While teaching the ‘Design of Experiments” (DOE) section of the Green and Black Belt exam syllabus I am often asked by students about the relevance of DOE to the service industry. Most students can see the obvious application of DOE to the Manufacturing sector however using DOE in service industries such as Banks, Call-Centre, Heath Service etc is not so obvious.
To help with this please see the White Paper published by Lou Johnson of MiniTab ( Minitab.com).
Lou Johnson describes a few great examples of successes using DOE in service industry. See examples below:
- A 15-factor supermarket test uncovered 6 changes that led to a 150% increase in product sales. (Direct Marketing, December 1997)
- A national carpet retailer increased their sales 20% by experimentally determining the sales person / customer interactions that optimized customer purchases. (Lean Six Sigma for Service, 2002)
- A 19-factor direct mail credit card test pinpointed 5 significant effects for a 15% jump in response rate.
(International Journal of Research in Marketing, 2006)
- GE Capital saved over $3MM by implementing the results of a 7 factor designed experiment studying methods to collect unpaid debt. (Quality Engineering, 2000)
- A global newspaper tested 11 creative and 4 price elements in one mail drop for a 41% increase in net response.
S(presented by Financial Times at DMA07)
The Conclusion drawn in the Whitepaper are also important and powerful:
Designed experiments have tremendous potential but are greatly under-utilized in the Service Quality industries. In particular, in-market testing – including retail, direct mail, Internet, and advertising testing – provides ample opportunity to leverage the experimental design and analysis techniques that have been developed in the past three decades. But the enormous potential comes with unique challenges. Far from textbook conditions, the “front lines” of marketing and service operations deal with the uncertainty of human behavior with textbook statistics often the first casualty. However, success is within reach. With the right statistical tools, good process knowledge, and a clear strategy, designed experiments can be completed, leading the way to improved sales and customer satisfaction.
