Under Control


Last April, I wrote a column entitled "Loose Logic for Fanatical Financiers," which seemed to strike a chord with a number of readers. I argued that failure leading to success, since research and development is susceptible to the laws of large numbers. A zero-risk endeavor cannot return value. Very high risks are prone to failure. We must therefor chose a risk or failure rate most likely to increase the value of our R& D investments. It's like this: a suggested failure rate for personal bank loans is two percent, buyers remorse for television hovers around 20 percent. Venture capitalists feel a failure rate of 80 percent is about right to obtain a total portfolio growth of 20 per year over a decade, based on the few projects that do succeed. Thus, a failure rate is chosen that leads to overall success. In like manner, the suggested target for R& D failure is about 50 percent. Choose each project for a 50/50 likelihood of failure. To pursue the "no fail" policy now in fashion will yield the equivalent of Czarist bonds in your stock portfolio. Small wonder large companies have little or no real output for the immense dollars spent on what they call R&D. Recently, I was fooling around with some software programs for formal evaluation of the process of evaluating R&D programs. It was suggested that the guidelines for the innovation segment ( about 10percent of the total evaluation ) should include eight bullet points. They are follows:

As Einstein is supposed to have said, "I can't help it if you don't believe it." Let's proceed to explain each of the mysterious Morley cues. 10X performance means that one parameter must be delivered with performance improved by a factor of ten. Not a soft metric, but hard. For example, the price the PC must be 10x lower or the performance must be 10x higher. Note that I said one parameter, not several. And several items of 10x dooms the project. Seven years implies a long time. The project must be managed into three-year segments over a five- to nine- year total life span. Nothing can work that targets the next quarter. Even engineering needed to sustain a product offering averages about 18 months' duration. I attend many a marketing meeting. Inevitably, the first question to the marketing people ask is "How long will the project take?" We have an answer: every project takes 18 months and a million dollars. WELL!! The meeting starts at 10 a.m. and we proceed to argue about the dollar amount and time needed. First, my partner and I concede that a 17-month schedule and a reduced budget is doable, but only because the target is the mighty "Dilbert" convention in Cleveland. By three in the afternoon we've agreed to finish up in 11 months, 4 days, and 7 hours with a budget of only $438,974.89. The marketing people smile and congratulate each other. But as my friend and I saunter down the hall, he turns and says to me, "Those silly tweeps actually think they're going to get the project done in 11 months." The truth is, we only agreed to the schedule, so we could break for lunch. Three "gurus" is the number of "smart" people needed for a project. These gurus need to be supported by no more than five persons. Additional people reduces the likelihood of success and increases the time and money involved. A large team needs to be broken up into small teams with clearly defined shippable deliverables. Every project, no matter how large, needs to be broken up into squad patrol size. Well, that's three of the eight parameters and we've already run out of space for this month. If you want to find out about the five-legged dog, the three rounds, and the importance of folding, you'll have to tune in next time.

As appeared in Manufacturing Systems Magazine January 1996 Page 84

Manufacturing Systems Magazine Articles
References - Table of Contents

Send mail to rmi.info@barn.org for more information.
Please send mail to webmaster@barn.org regarding web site structure.
Copyright © 1996-2002 R.Morley Inc. All Rights Reserved

R. Morley Incorporated
614 Nashua Street, Suite 56
Milford, NH 03055-4992 USA
Tel: 603-878-4365 FAX: 603-878-4385