Decision Tree Analysis for Making Decision Under Uncertainty

Share This Post

Share on facebook
Share on linkedin
Share on twitter
Share on email

Product Development:

Before discussing anything on product development let us discuss what is the product? Precisely, a product is ‘Any Thing or Things produced’. More precisely, as per engineering and manufacturing perspective, products are purchased as a raw material and sold as finished goods. In a real world, Product can be tangible or intangible.

Typically, either conceiving a design using CAD/CAM tools or reverse engineering is practiced as the best methods for a product development. Although these either conventional processes save time and cost to a launch product, but because of poor product feasibility and financial analysis product fails earlier before meeting the foresee financial gains. So what needs to be followed to develop the product? Are designers really following the correct path to develop the product? Do designers ever think to follow Product Life Cycle tools & techniques to achieve optimum solutions? What are the significance of following Product Life Cycle phases in terms of cost, efficiency, quality, reliability, durability, productivity and much more? It requires a lengthy topic to cover all answers to these questions. In brief Product life Cycle and its significance are described in the next section.

Product Life Cycle:

Traditionally, it is deemed to visualize the product life cycle with engineering perspective only as shown in figure-1:

Free trade, rapid product evolution and increase demand for a cheap product have changed the way to product development concept and visualization. In a broader spectrum, product Life Cycle is a process from product initialization to product sales & marketing.  Even today it covers CRM (Customer Relationship Management).

Product Life Cycle covers:

  • Introduction
  • Growth
  • Maturity (Where Product gets maximum penetration in the market)
  • Decline

Decline phase is also referred to as the evolution phase where the shift of technology or advancement in existing product may occur. For example,VCR was very popular during 90’s and now it has been replaced by the DVD player.

 

Pic 1 - Product Life Cycle
Pic 2 - Product Life Cycle

Decision Tree:

Decision Tree Analysis is widely used for product selection, alternative process selection and cost-benefit outcomes, etc. Decision Tree is one of the best quantitative decision-making methodologies. It helps to calculate the Expected Monetary Value (EMV).

                                                                           EMV = P X I

(Where P refers to probability and I refers to Impact)

Decision Tree is required to construct from left to right with all possible alternatives. Then put all the possible outcomes of each alternative. Decision and chance node are a constituent part of it. It is important to know that the sum of the probability of each alternative at a chance node should not be greater than 100% of 1 (0-1). The figure given below illustrates a basic concept of a Decision Tree.

 

 

blank
Pic 3 - Basics of Decision Tree

Decision Tree Analysis for Decision Making – Case Scenario:

In context to an earlier discussion, the right decision is important to achieve product objectives in a shorter time with minimum efforts. So I am going to discuss a real case scenario.

Company XYZ* decided to setup a manufacturing and assembling plant. They had available capital of more than $ 110 million. One of the directors of the company was a bit shaky to invest this amount because of the economic crunch, inflation, rapid trend changes, variety of similar products and decreasing buying capacity of the customers.

Therefore, company Top management decided to have an alternative solution and pick the best one on the basis of Quantitative Risk Analysis. It was decided to go with Decision Tree analysis along with other Quantitative Risk Analysis methodologies.

After initial assessment and research following data was obtained.

Data Sheet:**

  • Production:

Units/Day                    =                      3,000

Units/Month                =                      78,000

Units/Year                   =                      936,000

Units for 5 Year           =                      4,680,000

  • Cost***

Unit Price                    =                      $ 40

Unit Price excluding GST

(Ex Factory)                 =                      $ 25.70

Unit Cost                     =                      $ 13.70

Unit Profit                   =                      $ 12

  •  Internal Rate of Return (IRR) and Net Present Value (NPV)

 

As narrated earlier, factors that were taken-in-account for making right decision:

  • Economic Conditions
  • Capital In hand
  • Cash Flow
  • Loan Required
  • Interest Rate
  • Market Condition
  • Competitors etc.
blank
Table-1: Setup Cost

Development of Decision Tree:

After thorough analysis, there were five proposed options to choose the optimum one. Options were:

  • Complete Product Development along with Sales and Marketing
  • Product Design
  • Prototyping
  • Production
  • Sales and Marketing

Solution one was a complete product development from product initialization to sales and marketing. Other options were product design, prototyping, production and Sales and Marketing respectively. The goal was to get maximum profit by spending less capital while catering other internal and external Constraints as described in the last paragraph of section-4. Collapsed Decision Tree is illustrated in figure 4.

Incomplete product development, a risk of financial loss was greater because of substantial capital as well as production costs.

Development of Decision Tree

                                                  Figure-4: Development of Decision Tree

Complete Decision tree

                                                   Figure-5: Complete Illustrated Decision Tree

 

Outcome of Decision Tree:

After analyzing Decision Tree got the following results. Table -2 illustrates Arrival Probability (This is the probability that this decision encountered, assuming you followed all the optimal choices.) and Benefit of Correct choice (This is the difference of expected values between the optimal choice and the worst choice for this decision.).

Outcome of Decision Tree:

After analyzing Decision Tree got the following results. Table -2 illustrates Arrival Probability (This is the probability that this decision encountered, assuming you followed all the optimal choices.) and Benefit of Correct choice (This is the difference of expected values between the optimal choice and the worst choice for this decision.).

Optimal Choice

Arrival ProbabilityBenefit of Correct Choice
Sales & Marketing100.0000%

     $ 31,173,826.80

                                                                   Table-2: Decision Table

 

 

blank
Figure-6: Optimal Tree

‘Probability Chart’ and ‘Optimal Tree’ clearly supports the decision. The probability chart indicates the probability of the possible outcome of any decision. Figure-7 indicates that outcome for complete development and Sales and marketing in major. Graphically, it is evident, a possibility of financial loss with low and normal (medium) efforts to market and sell the Juicer Blender while aggressive generation revenue of more than $ 30 million in six years with marketing and sales only.

blank
Figure-7: Probabilities for Decision Tree
blank
Figure-8: Cumulative Probabilities for Decision Tree

Conclusion:

Expected Monetary Value (EMV) helped to select optimal part of Decision Tree. By sticking to Sales and Marketing only, the risk to financial loss minimized with a great opportunity to concentrate on one area only.

*Name of the company is not displayed as per request.

** Only shown relevant data.

 

Thank You

blank

@RISK

blank

More To Explore

Leave a Reply

Your email address will not be published. Required fields are marked *