Scribd
Upload
57 views

Decision Analysis: Saurabh Chandra

Tate was deciding whether to invest $300,000 in developing a new ergonomic chair. There was uncertainty around future market demand. A decision tree was used to analyze the options. It showed that with market research costing $50,000, the expected value of investing was $150,000. Without research, the expected value was $80,000. Therefore, the decision tree analysis indicated Tate should invest in market research before deciding to invest in chair development.

Uploaded by

Kunal Kaushal
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
57 views

Decision Analysis: Saurabh Chandra

Tate was deciding whether to invest $300,000 in developing a new ergonomic chair. There was uncertainty around future market demand. A decision tree was used to analyze the options. It showed that with market research costing $50,000, the expected value of investing was $150,000. Without research, the expected value was $80,000. Therefore, the decision tree analysis indicated Tate should invest in market research before deciding to invest in chair development.

Uploaded by

Kunal Kaushal
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 17

DECISION ANALYSIS

Saurabh Chandra

Decision Trees
Businesses involve uncertainty in strategic/tactical decisions
We discuss a set of techniques to analyze decisions under
uncertainty
The class of decisions have three properties:
Alternatives are well defined
The critical uncertainties can be quantified
The objectives are clear

Most of the decisions can be modified/presented in this form


Decision Trees are graphical devices:
Help to identify best option
Value of the alternatives

Alison Tate:
Manufactures low cost ergonomic chairs sold online
Popular modes with annual sales of $200,000 to $450,000
R&D develops new improved chair prototype
Investment required $300,000
Expected market demand:
Good: 8000 chairs x $100 = $800,000
Poor: 1000 chairs x $100 = $100,000

Need a way to determine the value of uncertain prospects


Probability tree way to evaluate uncertain gains and losses/value of
investment
Decision tree sequence of decisions

Probability Trees:
Laying out all the options:
Invest and large market
Invest and small market

Quantifying each option and choosing the decision if


odds are favorable
Invest and market
favorable

800

How do we measure
uncertainty?
How to assign probabilities?

Invest and market


unfavorable

100

Decision Trees:
800

Invest and market


unfavorable

100

0. =
4

Invest and market


favorable

p
0. =
6

Expected monitory value of the investment net of costs =


0.4 * $800,000 + 0.6 * $100,000 - 1.0 * $300,000 =
$80,000
- Probability weighted average of the outcomes

More complicated trees:


Chances are that the chairs may be shown on TV
This will increase sales four fold.
$400,000 for small market
$3.2 million for large market

Chances of showing on TV:


If market is large - 10%
If market is small 1%

Prob*Outco
me

TV promotion

3200

0.04

128

No TV promotion

800

0.36

288

TV promotion

400

0.006

2.4

No TV promotion

100

0.594

59.4

0. =
1

Probabili
ty

market favorable

0. =
01

=
p
6
0.

market
unfavorable

0. =
4

p
0. =
9

p
0. =
99

Value of investment opportunity net of costs = 477.8 300 = $177,800

Decision trees

We did not make any decision yet, only the value of


expected outcome
In reality we may be able to partially control events as
time goes on
To analyze a series of decisions where there is
uncertainty.
Chronological sequence of events presented in a
graphical manner
Decision trees serve two primary goals:
Help
decide whichUncertain
decision to make
Decisi
Outcomes
events/
chance
on
Value of any decision or set of options
nodes

Tates ergonomic chairs; a simple


decision problem
Tate was faced with deciding whether to invest $300,000 in a new
manufacturing facility when she was unsure whether there would be
sufficient demand for the chair it would produce to cover the large
investment.
If large market 8000 chairs with profits of $100 each = $800,000
If small market 1000 chairs with profits of $100 each = $100,000
A market research firm submitted proposal to evaluate market
potential
Cost = $50,000

Designing the tree; Step 1 lay out


the alternatives
Research
favorable

Do not
[0]
Invest

market research

Research
unfavorable
market
favorable

No market
research
[0]

Do not
[0]
Invest

market
unfavorable
market
favorable
market
unfavorable

[800]

[0]

Invest
[300]

[-50]

Invest
[300]

Invest
[300]

Do not
[0]
Invest
[800]

[100]
[0]

[0]

[100]

[0]

Designing the tree; Step 2 Quantify theInvest


uncertainties
[Research
favorable
p=
0.4

market research
[-50]
Research
unfavorable
p=
0.6
D

No market
research
[0]

Invest
[300]

Do not
[0]
Invest

market
favorable
p=
0.4
p=
market
0.6
unfavorable
market
p=
favorable
0.4
p=
market
0.6
unfavorable

[800]

300]

Do not
[0]
Invest

[0]

Invest
[300]
Do not
[0]
Invest
[800]

[100]
[0]

[0]

[100]

[0]

Designing the tree; Step 3


Specify the objectives

Research
favorable
p=
0.4

market research
[-50]
Research
unfavorable
p=
0.6
D

No market
research
[0]

Invest
[300]

Do not
[0]
Invest

market
favorable
p=
0.4
p=
market
0.6
unfavorable
market
p=
favorable
0.4
p=
market
0.6
unfavorable

Invest
[300]
Do not
[0]
Invest

[800]

[0]

Invest
[300]
Do not
[0]
Invest
[800]

[100]
[0]

[0]

[100]

[0]

Solving the tree

Mark all the decision nodes; A, B, C, and D


Start from the right side of the tree and work backwards
Let suppose she is in position A:
What to do?

Invest
[300]
Do not
[0]
Invest

[800]

A
[500]

[0]

Investing would yield a cash flow of $500,000 as


compared to not investing.

[300]
Do not
[0]
Invest

[100]
B
[0]
[0]

Invest
[300]

Do not
[0]
Invest

market
favorable
p=
0.4
p=
market
0.6
unfavorable
market
p=
favorable
0.4
p=
market
0.6
unfavorable

[800]

[100]
[0]

[0]

C
[80]

Research
favorable
p=
0.4

A
[500]
First path:
500*0.4 + 0*0.6
50
= $150,000

market research
[-50]
Research
unfavorable
p=
0.6

B
[0]

No market
research
[0]

C
Second path:
[80] = $80,000

Whats the final


conclusion?

D
[150]

Canonical decision problems case:


Canonical reduced to simplest or clearest schema
possible
Vignette - a brief evocative description, account, or
episode
Presented simplified versions of some standard
problems in management

Problem 1: Product development sequencing

You might also like