Perspective Triangulation
Compare informed viewpoints to separate signal from self-interest
- Difficulty
- Easy
- Time to result
- ~days to results
- Steps
- 5
- Confidence
- 96%
Perspective Triangulation gathers comparable judgments from several informed people and adjusts for the incentives shaping each answer. Begin with a concrete question, then consult sources who occupy different positions in the system. Compare where their accounts converge, investigate material disagreements, and explicitly discount claims that advance a source's interests. The output is not a vote or an average; it is an independent assessment informed by multiple vantage points. Dudley used this approach during the financial crisis because his Wall Street contacts could explain fast-moving conditions, but their views sometimes reflected their own positions. Speaking with three or four people helped him distinguish market-wide signals from self-serving narratives and detect operational risks that aggregate data could not yet reveal.
Origin
Bill Dudley described using multiple Wall Street contacts to assess fast-moving conditions while leading the New York Fed during the financial crisis.
Core principles
- 01Every source has a perspective and potential bias
- 02No single informed opinion establishes the truth
- 03Disagreement between sources reveals useful uncertainty
- 04Operational knowledge improves interpretation
How to run it
- 1
Frame the decision
Define the specific uncertainty that must be resolved and the decision it affects. Ask a question concrete enough that different answers can be compared.
Pro tip Request observable facts and mechanisms, not just predictions.
Watch out A vague question produces opinions that cannot be meaningfully reconciled.
- 2
Select diverse sources
Choose three or four knowledgeable people with different roles, exposures, or incentives. Include participants close enough to observe the system directly.
Pro tip Map what each person gains or loses under different outcomes.
Watch out Several sources from the same institution may still represent only one perspective.
- 3
Gather comparable accounts
Ask each source substantially the same core questions. Capture their evidence, assumptions, and confidence rather than only their conclusions.
Pro tip Ask what evidence would change each source's mind.
Watch out Do not reveal earlier answers in ways that anchor later sources.
- 4
Adjust for incentives
Identify where a source may be talking their book or protecting an institution. Discount unsupported claims without automatically dismissing informed evidence.
Pro tip Separate factual observations from preferred interpretations.
Watch out Treating self-interest as proof of dishonesty can discard valuable information.
- 5
Synthesize independently
Use convergence, disagreement, and incentive analysis to form your own assessment. State remaining uncertainty and what should be monitored next.
Pro tip Give extra weight to independently corroborated operational facts.
Watch out Triangulation improves judgment but does not eliminate uncertainty.
In the wild
During the financial crisis, Dudley asked a market participant how long it would take to value a complicated CDO containing many mortgages and tranches. The answer—at least two or three weeks—revealed that institutions could not quickly establish what important assets were worth. That operational fact made the severity of the market problem clearer than a generic confidence survey would have.
→ The inquiry exposed a valuation bottleneck that could intensify financial stress.
A manufacturer hears that a critical supplier may fail. Its finance team, the supplier, a major customer, and an industry lender each provide different accounts. Comparing payment delays, order cancellations, and credit conditions while accounting for incentives reveals that the supplier has a temporary liquidity problem rather than an irreversible demand collapse.
→ Management builds a limited contingency plan without prematurely terminating the relationship.
Common mistakes
Counting votes
Triangulation is not majority rule. Several sources can repeat the same mistaken assumption or depend on the same underlying information.
Ignoring incentives
An expert's access does not make their interpretation neutral. Identify how their position may shape what they emphasize or omit.
Seeking only familiar voices
Consulting people from one network can create false confirmation because their information and incentives overlap.
Is it for you?
Best for
It is best for leaders evaluating complex situations where reliable public data is limited or delayed.
Not ideal for
It is unnecessary for routine decisions supported by complete, objective, and timely data.
From the transcript
“So you need to talk to three or four people to sort of triangulate and figure out what you think is really going on.”
“I could call up people I knew and just ask their opinion, recognizing that oftentimes their opinion does have a touch of self interest.”
From the episode
Bill Dudley on Monetary Policies
Bill Dudley