Beyond quality.

While there is a focus on quality in many organizations, Keki Bhote and Adi Bhote, both of whom are (a) consultants and (b) active participants in Motorola's Six Sigma program (with the former being one of the creators of the program), think that there is an insufficient amount of attention being paid to another aspect of products, an aspect that means a whole lot to customers: reliability. As they write in World Class Reliability: Using Multiple Environment Overstress Tests to Make it Happen (AMACOM; $39.59), "Reliability ... is little known, ill-defined, and at best, relegated to be a mere hand-maiden of quality." They note that many companies "take a cavalier attitude in dismissing their costs associated with reliability," to the extent that if warranty costs are 1-2% of sales, they think they're doing just fine but, the authors point out, direct warranty costs are just the proverbial "tip of the iceberg," with other costs--field service, engineering, investigation, retrofit, customer downtime, customer anti-referrals, etc.--multiplying that warranty cost by as much as 100.

The authors are promoting a technique called "multiple environment overstress tests," or MEOST, which they claim "can dramatically improve reliability right at the design stage, in hours, instead of waiting months and years to confirm unreliability in the field." They maintain that reliability is a design "responsibility, not a quality function" because design, they stress, has a tremendous effect on product cost and performance: they claim that 70-75% of cost and >90% of field failures are related to design.

Although the subject of the book is ostensibly MEOST--which involves testing a product, under various conditions and stresses, to failure--and while they do make a good case for how this can provide tremendous benefits to companies that use it (e.g., they show that in the case of a car door, traditional engineering testing would require 172 hours of test time, 117 man-days of labor, $124,000, multiple prototypes, and would result in the discovery of 1 failure mode, while MEOST requires 16 hours of test time, 38 man-days of labor, $44,000, and just a few prototypes)--what is perhaps more important is their step-by-step review of how managers and engineers who actually may think they're pursuing a high-quality path to development and production are actually wandering down non-productive paths. This is particularly true--and troubling--as management typically drives designers and engineers to reduce costs. The consequences of cost reduction can be more devastating to a company's bottom line than the deployment of "costlier" but effective components leading to products that will delight the customer. (MEOST is a means to quickly rooting out the parts that just don't cut it from a functional point of view; the delight part is, well, perhaps, magic.) As the old line has it: You get what you measure. But if you're using the wrong measurement tools, no matter how often you measure, you may not end up with what you want--GSV