Value of Information
Hubbard's Value of Information framework reveals that effective decision-making hinges on identifying what information could actually influence a decision and determining the most efficient way to acquire it, especially under uncertainty.
Douglas Hubbard's Value of Information framework cuts through a recurring question: Should we research first or decide now? Information is only valuable if it might actually change your decision, and the cost of getting it is less than the expected improvement in outcomes.
Hubbard frames this as three questions.
- Will this information change the decision? When a museum proceeds with a program redesign regardless of what visitor research shows, the research has zero value. The discipline is asking what the research would need to show to change the choice. If the answer is "nothing would change the decision," that's manufacturing justification, not doing research.
- What's the cost of getting it? The obvious cost is monetary—a $30K visitor research study, for instance. But museums often underestimate non-monetary costs: a six-month study that validates the decision arrives after the funding cycle has closed and the board has moved on to other priorities.
- How much does reducing uncertainty improve the outcome? A $100K gallery redesign could generate $200K in value if successful. At 50% confidence in the return, the expected value is $100K (0.5 × $200K). Research that raises confidence to 75% shifts the expected value to $150K — a $50K improvement. Does that justify $30K in research costs plus delay? The precision matters less than the discipline of quantifying uncertainty before committing resources.
A counterintuitive implication: when uncertainty is high, small measurements can shift decisions. Hubbard shows that if you know almost nothing, almost anything will tell you something. Museums often assume major uncertainty requires major studies — months of visitor research, comprehensive surveys, longitudinal tracking. But a $100K decision at 20% confidence might shift to 60% confidence after three phone calls to peer institutions and a weekend of gallery observation. The question isn't whether to research, but what would actually change the bet.
VOI shifts the question from "Should we do research?" to "What information would change this decision, and what's the most efficient way to get it?"
This connects to Strategic Decisions as Bets — if decisions are bets under uncertainty, VOI asks when you should pay to reduce that uncertainty before placing the bet.