Own the Bad Call
Admit a wrong prediction, preserve the research lesson, and rebuild trust.
- Difficulty
- Easy
- Time to result
- ~days to results
- Steps
- 6
- Confidence
- 96%
Own the Bad Call is an accountability method for responding when a public forecast, recommendation, or strategic judgment proves wrong. The practitioner begins with an unqualified admission rather than softening the result or rewriting the original position. The prior analysis is preserved so the team can distinguish a reasonable process with a failed assumption from weak or biased research. The error is then decomposed: identify the assumption, mechanism, data limitation, or timing judgment that failed and convert it into a concrete process change. Publish the updated view and move on rather than endlessly defending the past. Because the correction is visible and specific, audiences can judge both the original reasoning and the quality of the response, increasing trust in future calls.
Origin
Extracted from Masters in Business during Ellen Zentner's discussion of failure and accountability in finance.
Core principles
- 01A wrong result does not automatically mean the original process was dishonest.
- 02Clear admission builds more trust than revisionist history.
- 03Extract the lesson without rewriting what was previously said.
- 04Move forward with an updated process and thesis.
How to run it
- 1
Admit the result
Say directly that the call was wrong. Avoid euphemisms that obscure whether the predicted outcome occurred.
Pro tip Use one clear sentence before offering context.
Watch out Do not begin with excuses or blame external events.
- 2
Freeze the original record
Retain the forecast, evidence, assumptions, and publication date. Evaluate what was actually claimed rather than what the forecaster now remembers claiming.
Pro tip Use archived reports, recordings, or decision logs.
Watch out Editing the historical record destroys the trust this method is meant to rebuild.
- 3
Locate the failed assumption
Determine whether the error came from data, causal reasoning, timing, probability calibration, or an unforeseen event. Separate process quality from outcome luck.
Pro tip Ask which single assumption would have changed the conclusion if corrected.
Watch out Do not label every miss an unforeseeable shock.
- 4
Convert failure into a process change
Specify how future analysis will treat the failed assumption or missing evidence. Make the adjustment narrow enough to test.
Pro tip Add a trigger, scenario, or pre-mortem question rather than promising to be more careful.
Watch out Avoid overcorrecting the entire model because of one outcome.
- 5
Issue the revised view
Explain what the new evidence means and what conclusion now follows. Keep the correction distinct from the historical forecast.
Pro tip State what would cause the revised view to change again.
Watch out Do not hide the correction inside an unrelated report.
- 6
Move forward transparently
Apply the lesson to the next call and allow the audience to evaluate the improved process over time.
Pro tip Review corrected forecasts periodically to detect recurring error patterns.
Watch out Repeated admissions without process improvement become performative.
In the wild
An economist forecast a central-bank increase that did not occur. She publicly states that the call was wrong, preserves the analysis, and finds that she overweighted one official's comments while underweighting broader committee guidance. Future forecasts require a documented voting-distribution assessment.
→ Clients receive a usable correction and can see how the forecasting process improved.
A founder predicted strong demand in a new region, but adoption remains weak. The founder acknowledges the miss, reviews the decision log, discovers that early interviews came from unusually enthusiastic customers, and adds representative sampling to future market validation.
→ The failed launch becomes a specific research improvement rather than a hidden embarrassment.
Common mistakes
Practicing revisionist history
Claiming that the original prediction meant something different prevents honest diagnosis and undermines the credibility of later calls.
Confusing explanation with excuse
Context helps locate the failure, but it should follow a direct admission and lead to a concrete process change.
Learning only from outcomes
A sound probabilistic call can fail, while a poor process can get lucky. Review the quality of the reasoning as well as the result.
Is it for you?
Best for
Analysts, executives, and public experts whose predictions or recommendations can be audited later.
Not ideal for
Cases involving misconduct or concealed evidence, which require formal investigation rather than a simple correction.
From the transcript
“say that was a bad call. I thought I had sound research behind it. Turns out out was I was wrong. And then let's move…”
“People will trust you so much more when you make the next call.”
“I've met plenty that have tried to cover it over time, revisionist history.”
From the episode
Ellen Zentner's Shift From Public to Private Sector
Ellen Zentner