Primary navigation:

QFINANCE Quick Links
QFINANCE Reference

Home > Business Strategy Checklists > Understanding Decision-Tree Analysis

Business Strategy Checklists

Understanding Decision-Tree Analysis

Checklist Description

This checklist describes decision-tree analysis and how to create and evaluate a decision tree.

Back to top


In operational areas, a decision tree—also known as a tree diagram—is a tool for reaching decisions. It uses a diagram or model of decisions and their possible outcomes, including chance events, resource costs, and utility. A decision tree can be used to select the strategy most likely to attain a specific goal. Decision trees are also used as predictive models in data mining (the science of uncovering hidden patterns in data) and machine learning (the development of algorithms and other techniques that enable computers to “learn”).

Decision trees have three types of node:

  1. Decision nodes: In the diagram these are usually represented by squares;

  2. Chance nodes: Represented by circles;

  3. End nodes: Represented by triangles.

A tree is usually drawn from left to right, with splitting paths (burst nodes) but no converging paths (sink nodes). Thus, when drawn by hand, the diagram tends to get very big to the right.

Decision trees can be a very effective structure for exploring options and investigating the consequences of choices of action. They can also help to form a picture of the risks and rewards for each possible course of action. In a financial context, decision trees can help to determine the best strategies for investment.

Drawing a decision tree begins with the decision that needs to be made, usually represented by a small square on the left-hand side of a large sheet of paper. For each possible choice a line is drawn out to the right, with a short description written along each line. At the end of each line the result, or outcome, should be stated; this may be an uncertain outcome (circle) or another decision (square), and the result should be written above the symbol. The process is repeated as required from each new decision square, always annotated with descriptions. Once drawn, the tree should be reviewed, as it is unlikely that all possibilities will emerge during the first round.

To work out which option has the greatest value, the decision tree is evaluated by assigning a cash value to each possible outcome. For each circle (an uncertain outcome), the probability of each outcome is estimated as a percentage, with the total of all possible outcomes for each course of action equaling 100%. Obviously, best guesses are often required.

To calculate a tree value, starting on the right-hand side of the tree, each calculation is completed on reaching a node (square or circle) and then recording the result. To calculate the vale of an uncertain outcome (circle), the value of the outcomes is multiplied by the probability as previously estimated.

Back to top


  • Decision trees are simple to understand and interpret.

  • They are worth doing even with quite uncertain data. Intuitive insights can be gained based on descriptions of a situation by experts.

  • Decision trees lay out a problem clearly so that all options can be explored, and they allow a full analysis of the possible consequences of a decision.

  • They provide a method for quantifying the values of outcomes and their probabilities.

  • Decision trees assist in making decisions with existing information and best guesses.

  • Decision trees can be used to optimize an investment portfolio.

Back to top


  • Diagrams can become very large when drawn by hand.

  • Trees created from numeric datasets can be complex.

Back to top

Action Checklist

  • Identify the decision you need to make.

  • Draw a line to the right for each solution with a description.

  • Consider the outcome at the end of each line.

  • Repeat the process for each new decision.

  • When complete, review the tree, evaluate it, and calculate the values.

Back to top

Dos and Don’ts


  • Review a decision tree often.

  • Review the evaluation values regularly.


  • Don’t consider your first effort as final but continually review and revise it.

Back to top

Further reading


  • Cai, Jingfeng. Decision Tree Pruning Using Expert Knowledge: Cost-sensitive Pruning. Saarbrücken, Germany: VDM Verlag, 2008.
  • de Ville, Barry. Decision Trees for Business Intelligence and Data Mining Using SAS Enterprise Miner. Cary, NC: SAS Institute, 2006.
  • Rokach, Lior, and Oded Maimon. Data Mining with Decision Trees: Theory and Applications. Singapore: World Scientific Publishing, 2008.



Back to top

Share this page

  • Facebook
  • Twitter
  • LinkedIn
  • Bookmark and Share