Tool

Decision Value & Value-of-Information (VoI) Calculator

An interactive worksheet that captures action payoffs, state probabilities, analysis costs, and simple analysis-outcome scenarios. Saves your inputs and guides you through computing expected value and the value of information (VoI). Supports both a quick estimate and a two-outcome detailed VoI calculation.

Interactive Tool

Decision Value & Value-of-Information (VoI) Calculator

Purpose

Use this calculator to focus analytic effort where it will change decisions. Enter payoffs for two alternative actions under two possible states of the world, your current probability estimate, and the estimated cost of an analysis. Two calculation modes are supported: a quick estimate where you provide an estimated expected value after analysis, and a detailed two-outcome scenario where you provide likely analysis outcomes and their probabilities.

Key formulas (manual)

  • Expected value (EV) of an action = P(state1) * payoff_if_state1 + (1 - P(state1)) * payoff_if_state2
  • Current decision EV = max(EV(action A), EV(action B))
  • Value of Information (VoI), quick estimate = Estimated EV after analysis - Current decision EV - Cost of analysis
  • Value of (sample) information, detailed = Sum_over_outcomes[P(outcome_i) * EV(best action if outcome_i)] - Current decision EV - Cost of analysis

How to use this form

  1. Enter payoffs and your current probability to compute the current decision EV (shown in the example below).
  2. Use the quick estimate if you have a solid estimate of the expected improvement an analysis would produce (enter it as “Estimated EV after analysis”).
  3. Use the detailed two-outcome mode if you can think of two likely analysis outcomes (for example, the test says “high demand” or “low demand”), estimate their probabilities and the EV you would get under each outcome.
  4. If VoI >= cost of analysis (and the result justifies implementation effort and timing), prioritize the analysis; otherwise deprioritize.

Example

Marketing reallocation example (simplified): Action A = keep current spend, Action B = reallocate 20% to channel X. Two states: High response or Low response. Suppose you estimate P(High)=0.4. Action A EV = 0.4*100 + 0.6*80 = 88. Action B EV = 0.4*140 + 0.6*50 = 86. Current decision EV = max(88,86)=88 (choose A).

Quick estimate: you estimate analysis would let you achieve EV=95 on average; cost_of_analysis=3 → VoI = 95 - 88 - 3 = 4 (positive, so analysis is worthwhile).

Short name for this decision (e.g., Marketing reallocation Q3)
Describe the decision, time horizon, and any implementation constraints.
Name of the first action (e.g., Keep current spend)
Enter monetary or utility payoff if State 1 occurs (same units across payoffs).
Enter payoff for Action A in State 2.
Name of the second action (e.g., Reallocate 20% to channel X)
Enter payoff for Action B when State 1 occurs.
Enter payoff for Action B in State 2.
Enter as a decimal, e.g. 0.4 for 40%. If you prefer percentages, convert to decimals.
Direct cost plus any expected implementation delay costs you want to include here.
If you already have a credible estimate of the expected EV that analysis would enable, enter it here to compute a quick VoI estimate.
E.g., probability that analysis will indicate 'high demand'. Leave blank if using quick estimate.
The EV you expect to achieve if the analysis returns outcome 1 and you then take the best action for that outcome.
Probabilities should sum to ~1 across outcomes; use two outcomes for a simple sample information estimate.
The EV you expect to achieve if the analysis returns outcome 2 and you then take the best action for that outcome.
Record key assumptions, sources, who provided estimates, and risks or constraints on implementing the chosen action.
Yes saves the input so you can reference and compare later.
You can explore this tool now. Sign in or create an account to save your responses and return to them later.
Sign In to Save Create an Account
Make this tool part of your work

Save a personal copy, bring it to your team, or tailor the questions and workflow to fit what you are hungry to improve.

Member customization and team collaboration are coming soon.
Discuss Ask the Librarian

Discussion

Comments and conversation will live here.