Max More
5.0 out of 5 stars This book matters!
Reviewed in the United States 🇺🇸 on November 7, 2003
The way to succeed is to double your failure rate. That comment by Thomas Watson, Sr. is not among the innovators' words of wisdom in Stefan Thomke's densely informative exploration of technologies and processes of experimentation but it perfectly fits the message. Central to Thomke's message in this book is the idea that iterated experimentation through the use of models, prototypes, and computer simulations is the key to learning and innovation. Getting the key to fit in the lock of increased organizational innovation capability, however, takes some jiggling and struggling. Experimentation Matters details the technologies that can transform innovation but place just as much emphasis on the changes that must be made to business processes, organization, culture, incentives, and management. Thomke provides plenty of detailed illustrations of companies wrestling with these issues, and offers six principles revolving to help companies experiment early and often and to organize for rapid iteration.
The first part of the book explains in depth the reasons why experimentation matters for learning and innovation, and how new technologies are affecting the development of both products and services. Thomke shows how the rate of learning is influenced by several factors that affect the process and how it is managed: fidelity, cost, iteration time, capacity, sequential and parallel strategies, signal-to-noise ratio, and type of experiment. Beneath the bewildering diversity of approaches to innovation in different industries, Thomke uncovers six principles that can improve how experimentation occurs: Anticipate and exploit early information through front-loaded innovation processes; Experiment frequently but do not overload your organization; Integrate new and traditional technologies to unlock performance; Organize for rapid experimentation; Fail early and often but avoid "mistakes"; and Manage projects as experiments.
six principles to help companies experiment early and often and to organize for rapid iteration that can improve how experimentation occurs:
(1) Anticipate and exploit early information through front-loaded innovation processes;
(2) Experiment frequently but do not overload your organization;
(3) Integrate new and traditional technologies to unlock performance;
(4) Organize for rapid experimentation;
(5) Fail early and often but avoid "mistakes"; and
(6) Manage projects as experiments.
In the final chapter, Thomke looks at how some companies are "shifting the locus of experimentation" to customers as a way to create new value. This approach, sometimes referred to as "co-creation", not only raises productivity but helps fundamentally change the sorts of products and services that can be created.
Innovation toolkits given to customers need to enable them to iterate through the steps of experimentation, be user-friendly, contain libraries of useful, pretested and debugged components and modules, and they must contain information abut the capabilities and limitations of the production process. In addition to the development of a customer toolkit, Thomke adds four other steps for shifting experimentation and innovation to customers and, very importantly, notes how the creation and capture of value also shifts.
One great strength of Thomke's book is the attention given to the managerial and organizational challenges of implementing new technologies such as computer modeling and simulation and combinatorial and high-throughput testing.
As other writers have repeatedly emphasized - but many managers have not yet understood - new technologies *must* be introduced only in concert with revised business processes, structures, and management approaches.
Iterated experimentation helps learning by increasing the number of failures. But if incentives continue to punish failures, the new technologies will be underused or misused. Financial incentives, organizational culture, and management communications will have to change if experimenters are to feel free to fail at the most productive rate.
feel free to fail at the most productive rate.
The way to succeed is to double your failure rate.
(5) Fail early and often but avoid "mistakes"; and
Thomke illustrates and details the crucial role of organization, process, and management in realizing the potential of experimentation technologies with a range of illuminating cases. He devotes a chapter to these effects in the integrated circuit industry, examines the challenges faced by Bank of America in its bold service experimentation efforts, and shows how managers at Eli Lilly struggled with non-technological aspects of high-powered experimentation in the drug discovery process.
A study of experimentation in the auto industry, particularly at BMW, suggests several lessons regarding the reality of technology introduction: Technologies are limited by the processes and people that use them; organizational interfaces can get in the way of experimentation; and technologies change faster than behavior. Thomke also shows how managers can look at projects as experiments, reiterating, refining, and learning from them as they proceed through the stages of design, build, run, and analyze.
B.Sudhakar Shenoy
5.0 out of 5 stars Innovation redefined
Reviewed in the United States 🇺🇸 on November 27, 2003
Observation, exploration and experimentation have been the three basic means of learning for scientists. Of these, experimentation calls for the highest levels of external intervention and as a topic by itself has always been of interest to statisticians who have developed powerful techniques to derive maximum information through the least possible number of experiments. Application of these statistical techniques has resulted in substantial reduction in research expenditure, quicker understanding of scientific principles and shorter time to convert ideas into useful products. On the other hand new technologies like simulation, CAD/CAE that harness the advances in computing have completely changed the experimental landscape by providing powerful techniques for rapid and economical experimentation on our desktops and servers. To cite one example discussed in this book, car maker BMW's crash simulation test progressed from 3000 to 700000 finite elements between 1982 to 2002 while simultaneously resulting in reduction of processing time from 3 months to 30 hours. Power of computing enables "front-loaded" innovation - understanding the phenomenon before committing resources into physical manufacturing.
But the lacuna [a space where something has been omited or has come out; gap; hiatus; esp., a missing portion in a manuscript, text, etc.] is that experimentation has never been thought as a separate management discipline cutting across functional silos to bring innovative solutions into the marketplace. Experimentation as a strategic tool that needs management attention and involvement is the core theme of this book.
Management deals with producing results under uncertainty. Uncertainty can be broadly classified under technical, production, market and customer needs. Experimentation should tell us not only what will work, but also what does NOT work. The knowledge so derived should seamlessly flow across the Design-Build-Run-Analyze cycle that cuts across departmental boundaries in large organizations. This is analogous to the concept of ERP in business processes. Though this concepts looks simple, organizational barriers prevent the seamless sharing of information for innovation. Design, manufacturing , marketing and procurement functions fail to optimize on the organizational repository of knowledge that can put winning products into the marketplace. This book is an excellent study on how management can use experimentation as a unique strategy within and beyond organizational boundaries. Case studies are quite detailed and well illustrated.
Read this book. It is worth experimenting.
No comments:
Post a Comment