Decision Tree
- 2. What’s a Decision Tree? A decision tree is a type of diagram that clearly defines potential outcomes for a collection of related choices. In project management, a decision tree analysis exercise will allow project leaders to easily compare different courses of action against each other and evaluate the risks, probabilities of success, and potential benefits associated with each.
- 3. It’s important to note that a proper decision tree has four main elements: Decision Nodes: A decision node, represented on our decision tree diagram as a square, indicates a choice that needs to be made. Chance Nodes: A circle represents a chance node and is used to signify uncertain outcomes. These nodes are used when future results are not guaranteed. End Nodes: End nodes, like the name suggests, represent the end of a diagram and illustrates a final outcome. Branches: Lastly, we have branches. Branches are what connect the nodes together. Each branch represents a potential choice and should be clearly labeled.
- 4. Example: Suppose a organization is using a legacy software. Some influential stakeholders believe that by upgrading this software the organization can save millions, while others feel that staying with the legacy software is the safest option, even though it is not meeting the current company needs. The stakeholders supporting the upgrade of the software are further split into two factions: those that support buying the new software and those that support building the new software in-house.
- 5. Building the Decision Tree: Build the new software: To build the new software, the associated cost is ₹50,000. Buy the new software: To buy the new software, the associated cost is ₹75,000. Stay with the legacy software: If the company decides to stay with the legacy software, the associated cost is mainly maintenance and will amount to ₹10,000. Stay/Buy/Build Decision Build the New Software Cost:₹50,000 Stay With the Legacy Software Cost:₹10,000 Buy the New Software Cost:₹75,000
- 6. The Buy the New Software and Build the New Software options will lead to either a successful deployment or an unsuccessful one. If the deployment is successful then the impact is zero, because the risk will not have materialized. However, if the deployment is unsuccessful, then the risk will materialize and the impact is ₹2lakh. The Stay with the Legacy Software option will lead to only one impact, which is ₹2lakh, because the legacy software is not currently meeting the needs of the company. Nor, will it meet the needs should there be growth.
- 7. Stay/Buy/Build Decision Build the New Software Cost:₹50,000 Successful deployment Cost:₹0 Unsuccessful deployment Impact:₹2lakh Stay With the Legacy Software Cost:₹10,000 Growth in Business Impact: ₹2lakh Buy the New Software Cost:₹75,000 Successful deployment Cost:₹0 Unsuccessful deployment Impact:₹2lakh 100% 40% 5%
- 8. Calculating Expected Monetary Value for each Decision Tree Path The diagram depicts the decision tree. Now, you can calculate the Expected Monetary Value for each decision. The Expected Monetary Value associated with each risk is calculated by multiplying the probability of the risk with the impact.
- 9. Build the new software: ₹ 2,00,000 * 0.4 = ₹ 80,000 Buy the new software: ₹ 2,00,000 * 0.05 = ₹ 10,000 Staying with the legacy software: ₹ 2,00,000 * 1 = ₹ 2,00,000 Build the new software: ₹ 50,000 + ₹ 80,000 = ₹ 1,30,000 Buy the new software: ₹ 75,000 + ₹ 10,000 = ₹ 85,000 Staying with the legacy software: ₹ 10,000 + ₹ 2,00,000 = ₹ 2,10,000 Now, add the setup costs to each Expected Monetary Value:
- 10. Conclusion Looking at the Expected Monetary Values computed in this Decision Trees example, you can see that buying the new software is actually the most cost efficient option, even though its initial setup cost is the highest. Staying with the legacy software is by far the most expensive option.