Decision Tree Analysis, Explained
Decision Tree Analysis
Plain-Language Definition
Decision tree analysis is a method for choosing between several options when the outcome of each option is uncertain but can be assigned probabilities and payoffs. It maps out every alternative, every possible condition affecting each alternative, and the resulting payoff — then calculates which choice produces the best result on average.
Why It Matters
Business decisions rarely come with guaranteed outcomes. Decision tree analysis gives you a structured, defensible way to choose under genuine uncertainty, rather than picking based on gut feeling or the most optimistic scenario.
The Structure
A decision tree has three core components:
- Decision node (usually drawn as a square) — the point where you choose between alternatives
- State-of-nature node (usually drawn as a circle) — the point where an uncertain condition (like market demand) plays out, with a probability attached to each possible state
- Payoff — the outcome value at the end of each branch, usually calculated from demand and profit figures
The Process
- Identify every alternative you’re choosing between
- For each alternative, identify the possible uncertain conditions and their probabilities
- Calculate the payoff for each combination of alternative and condition
- Calculate the expected value for each alternative (see: Expected Value: How to Calculate and Interpret It)
- Choose the alternative with the highest expected value — unless other factors (risk tolerance, strategic fit) argue otherwise
Worked Example (Fictitious Company)
Alderbrook Coffee Roasters is deciding between developing a new cold brew product, expanding its existing roast line, or maintaining current operations — each facing uncertain “high” or “low” market demand. Building the tree means mapping all three alternatives, their associated probabilities and payoffs, and calculating which produces the highest expected value (see the full worked calculation in the Expected Value explainer).
Key Takeaways
- Decision tree analysis is built for choosing between discrete alternatives under quantifiable uncertainty
- The technique is only as reliable as its inputs — probability and payoff estimates that are off will shift the recommended decision
- Decision trees are known to be sensitive to small changes in input data, a documented limitation worth naming in any analysis using this technique
- The highest expected value isn’t automatically the “right” choice — it’s the mathematically optimal choice given the stated assumptions, which is a narrower claim
Related Content
- C207 Course Roadmap
- C207 QUM3 Task 2 Guide: Decision Tree Analysis
- Expected Value: How to Calculate and Interpret It
References & Further Reading
- Lee, C. S., Cheang, P. Y. S., & Moslehpour, M. (2022). Predictive Analytics in Business Analytics: Decision Tree. Advances in Decision Sciences, 26(1), 1–29. — A peer-reviewed paper on decision tree methodology in business analytics, including its structural strengths and its sensitivity to input variation.