Setback Autopsy
Convert disappointing outcomes into specific operating improvements
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 6
- Confidence
- 95%
Strong performance attracts praise and can prevent decision-makers from hearing what they are doing wrong. A setback interrupts that feedback distortion and exposes assumptions, positioning choices, and operating weaknesses that success may have concealed. The method is to resist retreating into a shell, reconstruct the decision as it looked beforehand, distinguish uncontrollable conditions from controllable errors, and examine why the chosen response failed. The result should be a specific correction that can be applied to the next decision rather than a vague promise to be more careful. Samberg’s early fund loss during the 1987 crash illustrates the mechanism: illiquid small-company positions were especially vulnerable, and anticipated redemptions forced sales in a market without natural buyers. The useful lesson came from analyzing the interaction of strategy, liquidity, and forced execution.
Origin
Extracted from Masters in Business after Art Samberg described learning from his fund’s losses during the 1987 crash.
Core principles
- 01Success can conceal active mistakes
- 02Disappointment creates unusually honest evidence
- 03Analyze causes rather than retreating emotionally
- 04Improvement requires changing future behavior
How to run it
- 1
Record the outcome
Document what happened using contemporaneous facts, including the size, timing, and consequences of the setback.
Pro tip Write the account before hindsight reshapes the original reasoning.
Watch out Do not begin by assigning blame.
- 2
Reconstruct the decision
List the assumptions, available evidence, constraints, and intended outcome at the moment the decision was made.
Pro tip Compare the original thesis with what actually occurred.
Watch out Judging the decision only by its outcome produces hindsight bias.
- 3
Separate causes
Distinguish external events from exposure, preparation, execution, and other controllable factors.
Pro tip Ask which damage would still have occurred under a better process.
Watch out Calling everything bad luck prevents learning.
- 4
Identify the failure mechanism
Explain the causal chain that converted an incorrect assumption or weak preparation into the disappointing result.
Pro tip Look for interacting risks such as leverage, liquidity, and forced timing.
Watch out A list of symptoms is not a root-cause analysis.
- 5
Install a correction
Turn the lesson into a position limit, checklist item, review trigger, or other observable behavior.
Pro tip Assign the correction to a real future decision point.
Watch out Insight without a process change rarely alters performance.
- 6
Test and retain the lesson
Apply the correction in subsequent decisions and verify that it improves resilience without eliminating productive risk.
Pro tip Review whether the rule works across more than one incident.
Watch out Overcorrecting after one setback can create a new failure mode.
In the wild
Samberg entered the crash holding overlooked small companies that could become larger businesses. When those illiquid positions were hit and he anticipated investor redemptions, he had to call companies and ask whether they would buy his stock because other buyers had disappeared.
→ The loss exposed the connection between attractive small-company holdings, liquidity risk, and forced selling during market stress.
A team misses its launch target and initially blames weak demand. Its autopsy finds that customer interviews were conducted with enthusiastic early adopters rather than ordinary buyers, producing a distorted forecast. The team changes its validation checklist and tests the next offer with a representative sample.
→ A concrete research correction replaces a vague commitment to improve forecasting.
Common mistakes
Retreating from the evidence
Emotionally withdrawing after failure prevents the careful self-analysis needed to improve the process.
Treating success as proof of perfection
Positive outcomes can coexist with serious mistakes, especially when favorable markets conceal weak decisions.
Learning only a slogan
A lesson such as “take less risk” is too vague unless translated into a specific decision rule or control.
Is it for you?
Best for
It is best for investors, operators, and leaders making consequential decisions under uncertainty.
Not ideal for
It is not ideal when the result is too recent for objective review or when reliable evidence is unavailable.
From the transcript
“But you are doing things wrong.”
“really look carefully at what you did and why it failed”
“you will not do well.”
From the episode
An Interview With Arthur Samberg: Masters in Business (Audio)
Arthur Samberg