Utility Analysis: Fourteen Steps (Multiple Classes of Outcomes)
A step-by-step framework for evaluating options across multiple outcome classes by assigning utilities and probabilities, weighting each class, and computing total weighted expected values to rank options.
THE IDEA
The Fourteen Steps of Utility Analysis (Multiple Classes of Outcomes)
Step 1: Identify the options to be analyzed.
Step 2: Identify the outcome classes and each one’s range of outcomes and weight each class according to its importance.
Step 3: Identify the perspective of the analysis.
Step 4: Construct for each class of outcomes a utility matrix (Figure 16-14) with identical perspective and options.
Perform Steps 5 through 8 with each matrix.
Step 5: Assign utilities from 0 to 100 to the outcome of each option-outcome combination (each cell of the matrix). There must be at least one 100.
Step 6: Assign a probability to the outcome of each option-outcome combination (each cell).
Step 7: Compute expected values.
Step 8: Add expected values for each option and enter the totals in the “Total EV” column.
Step 9: Construct a single “merged” matrix (Figure 16-15) with the same options as in the classes-of-outcomes matrices.
Step 10: Enter opposite each option the total expected values for that option from the classes-of-outcomes matrices.
Step 11: Multiply the total expected values under each class of outcomes by the class’s weight.
Step 12: Add the resulting products (weighted expected values) for each option and enter the sums in the “Total Weighted EV” column.
Step 13: Rank the options. The one with the greatest total weighted expected value is the preferred option.
Step 14: Perform a sanity check.
The referenced matrices are not included here.
Additional note: The perspective in each matrix should be the owner of the problem—the person or entity that weights the outcome classes and must ultimately choose. Combining multiple-outcome and multiple-perspective analysis is discouraged as it may overcomplicate the analysis.
When to use it
Choosing between business strategies with multiple objectives (e.g., profit, market share, sustainability).
Selecting a job offer considering salary, location, growth opportunities, and work-life balance.
Evaluating policy options with social, economic, and environmental impacts.
Personal decision like buying a house considering cost, location, size, and condition.
FROM THEORY TO EVERYDAY LIFE
In practice
Investment Decision
A company is deciding between three projects. They evaluate each on financial return, risk level, and alignment with corporate values. They assign utilities and probabilities for each outcome under each class, weight the classes, and compute total weighted expected values to rank the projects under those assumptions, then check how changes in the inputs affect the ranking.
A MOMENT TO REFLECT
Put it to work.
How do you determine the weights for each outcome class in a real-world decision?
What are the challenges in assigning utilities to outcomes when the outcomes are not easily quantifiable?
How can you validate the probabilities assigned to outcomes?
CONTINUE THE CONVERSATION
Source texts
- The Thinkers Toolkit by Morgan D Jones