Page 560 -

540 CHAPTER 13 DECISION ANALYSIS

MANAGEMENT SCIENCE IN ACTION

Decision Analysis at Eastman Kodak
lemen and Kwit conducted a study to deter- decision analysis returned substantial value to the
C mine the value of decision analysis at the company. Indeed, they concluded that the value
Eastman Kodak company. The study involved an added by the projects was at least 185 times the
analysis of 178 decision analysis projects over the cost of the analysts’ time.
ten-year period from 1990 to 1999. The projects In addition to the monetary benefits, the authors
involved a variety of applications including strategy point out that decision analysis adds value by facil-
development, vendor selection, process analysis, itating discussion among stakeholders, promoting
new-product brainstorming, product-portfolio selec- careful thinking about strategies, providing a com-
tion and emission-reduction analysis. These projects mon language for discussing the elements of a
required 14 372 hours of analyst time and the decision problem and speeding implementation by
involvement of many other individuals at Kodak over helping to build consensus among decision makers.
the ten-year period. The shortest projects took less In commenting on the value of decision analysis at
than 20 hours, and the longest projects took almost Eastman Kodak, Nancy L. S. Sousa said, ‘As Gen-
a year to complete. eral Manager, New Businesses, VP Health Imaging,
Most decision analysis projects are one-time Eastman Kodak, I encourage all of the business
activities, which makes it difficult to measure the planners to use the decision and risk principles and
value added to the corporation. Clemen and Kwit processes as part of evaluating new business
used detailed records that were available and some opportunities. The processes have clearly led to bet-
innovative approaches to develop estimates of the ter decisions about entry and exit of businesses’.
incremental dollar value generated by the decision Although measuring the value of a particular deci-
analysis projects. Their conservative estimate of the sion analysis project can be difficult, it would be
average value per project was $6.65 million and their difficult to dispute the success that decision analysis
optimistic estimate of the average value per project had at Kodak.
was $16.35 million. Their analysis led to the conclu-
Based on Robert T. Clemen and Robert C. Kwit, ‘The Value of Decision
sion that all projects taken together added more
Analysis at Eastman Kodak Company, 1990–1999’, Interfaces (Sep-
than $1 billion in value to Eastman Kodak. Using tember/October 2001): 74–92.
these estimates, Clemen and Kwit concluded that

As we have seen in other chapters, managers are typically faced with uncertainty
about the consequences of their decisions. Typically, a manager will be faced with
several decision alternatives – the different decisions they can make in some sit-
uation. The difficulty arises because the outcomes from these alternative decisions
will to some extent be uncertain. We could decide to launch a new product but
future sales will be uncertain. We could decide to invest in the latest production
equipment to boost productivity but the financial benefits will be uncertain. A
hospital could decide to introduce the latest medical scanning facilities but the exact
effect on patients and patient throughput will be uncertain. In such situations, a
good analysis will include risk assessment where the risk associated with some
decision is the direct result of the uncertainty surrounding the outcome from that
decision. Decision analysis is a technique that can be used to help a manager
determine a suitable decision under such uncertain conditions.
We begin the study of decision analysis by considering problems having reason-
ably few decision alternatives and reasonably few possible future events. Payoff
tables are introduced to provide a structure for the decision problem and to
illustrate the fundamentals of decision analysis. We then introduce decision trees
to show the sequential nature of decision problems. Decision trees are used to

Copyright 2014 Cengage Learning. All Rights Reserved. May not be copied, scanned, or duplicated, in whole or in part. Due to electronic rights, some third party content may be suppressed from the eBook and/or eChapter(s). Editorial review has
deemed that any suppressed content does not materially affect the overall learning experience. Cengage Learning reserves the right to remove additional content at any time if subsequent rights restrictions require it.