Process-First Outcome Review
Judge decisions by the reasoning available at the time, not by their results
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 5
- Confidence
- 96%
Process-First Outcome Review separates the quality of a decision from the result that happened afterward. Before acting, the decision-maker records the thesis, evidence, assumptions, probabilities, and expected range of outcomes. Once results arrive, the review compares the original reasoning with what actually occurred without allowing hindsight to rewrite the thesis. A profitable investment can be classified as a process failure if it succeeded for an unanticipated reason, while a loss can remain a defensible decision when it fell inside the expected distribution. The review identifies whether skill, randomness, omitted evidence, or flawed execution drove the outcome. Lessons are then applied to future decision rules rather than used merely to praise winners or condemn losers.
Origin
Extracted from Masters in Business during Josh Wolfe's discussion of business failure, uncertainty, and the distinction between investment process and outcome.
Core principles
- 01Separate decision quality from outcome quality
- 02Evaluate reasoning using information available at the time
- 03Treat lucky profits as possible failures
- 04Expect some losses within a sound probabilistic process
- 05Improve the process instead of rationalizing the result
How to run it
- 1
Write the pre-decision record
Document the thesis, available evidence, critical assumptions, expected probabilities, and reasons for acting before the outcome is known.
Pro tip Include evidence that would change your mind.
Watch out A thesis written after the result is vulnerable to hindsight bias.
- 2
Define the expected distribution
Describe plausible positive, neutral, and negative outcomes rather than predicting a single point result. Note which losses would be normal consequences of uncertainty.
Pro tip Assign rough probabilities even when precise estimates are impossible.
Watch out Do not equate every negative outcome with an avoidable error.
- 3
Compare thesis with reality
After sufficient evidence arrives, assess whether events unfolded for the reasons originally expected. Separate thesis accuracy from financial or operational success.
Pro tip Use the original memo verbatim during the review.
Watch out Do not retrofit the thesis to match what happened.
- 4
Attribute skill and luck
Identify which parts of the result arose from analysis, execution, external randomness, or an unforeseen event. Classify lucky wins and unlucky losses explicitly.
Pro tip Ask whether the same process repeated many times would remain attractive.
Watch out Confidence should not increase merely because one favorable outcome occurred.
- 5
Update the process
Convert the review into changes to research, sizing, decision rules, or execution. Preserve sound practices even when one outcome was unfavorable.
Pro tip Create one concrete process change for each material error.
Watch out Avoid broad lessons based on a single noisy case.
In the wild
A fund invests because it expects a startup's original product to dominate a specific market. The product fails, but an unrelated buyer acquires the company for its engineering team, producing a gain. The fund records the financial result as positive but classifies the original decision process as flawed because the expected mechanism never occurred.
→ The team avoids reinforcing a weak thesis simply because luck produced a profit.
Common mistakes
Equating profit with correctness
A favorable outcome can result from luck or from a mechanism absent from the original thesis.
Punishing expected losses
Some sound probabilistic decisions will lose even when the process is working properly.
Rewriting the thesis
Changing the remembered rationale after results arrive destroys the value of the review.
Is it for you?
Best for
It is best for investors, executives, and teams making repeated decisions under uncertainty.
Not ideal for
It is not ideal for deterministic work where a correct procedure should reliably produce one specific result.
From the transcript
“We watched as process first outcome right”
“So there are times where we actually make money, but we considered it a failure because we have the wrong process.”
“We ended up making money, but we're totally wrong. We consider that a process failure.”
From the episode
Josh Wolfe Discusses Innovative Investments (Podcast)
Josh Wolfe