A contractor arrives late and the project slips. After the result is known, the original choice can sound obviously wrong. But was the decision poor given the information available, or did a reasonable choice meet a risk that actually materialized?
Record consequential decisions before their outcomes arrive. Include the options, evidence, uncertainty, and assumption that would reverse the choice. Later, review the process separately from the result. A decision journal helps preserve context; this particular tool is not a research-validated intervention.
The outcome is information, not the whole evaluation
Baron and Hershey's 1988 studies found that knowledge of favorable or unfavorable outcomes influenced people's evaluations of decisions made under uncertainty. Their participants evaluated described medical decisions or monetary gambles. The findings do not establish that a workplace journal eliminates that bias. See the original outcome-bias paper.
Fischhoff's 1975 experiments separately examined how knowing an outcome changed judgments about what had been foreseeable. Participants were not fully aware of that influence. See the original hindsight research. These are related but different problems: judging the choice by its result, and reconstructing what could have been known beforehand.
A dated record does not make the writer objective. It gives the review something more concrete than a confident memory. Keep the original entry intact and add later observations separately.
Record a choice with a genuine alternative
Illustrative scenario: Anika needs specialist capacity for a delivery milestone. She can hire a contractor with relevant experience or train two internal engineers while reducing near-term scope. The contractor is available sooner in principle, but the start date depends on access approval. Training creates longer-term capability but cannot cover the full immediate commitment.
Anika chooses the contractor, conditional on confirmed access and a bounded initial task. She records that the sponsor values the immediate milestone more than broader scope, that internal training continues in parallel, and that an unconfirmed start date would reverse the choice.
She does not write “contractors are faster” as a universal fact. She names the evidence for this person, the dependency, and the consequence. If access is not confirmed by the agreed checkpoint, she will reduce scope and proceed with the internal plan.
The journal's useful part is the condition. It prevents a tentative assumption from becoming an invisible certainty after the decision is announced.
A record short enough to use
| Field | Filled example |
|---|---|
| Decision and date | Use specialist contractor for the bounded milestone task. Record before committing. |
| Options | Contractor; internal training with reduced immediate scope. |
| Evidence available now | Relevant prior work reviewed; task defined; internal capacity constrained. |
| Important uncertainty | Access approval and confirmed start date. |
| Tradeoff accepted | External dependency in exchange for near-term specialist capacity. |
| Reversal condition | No confirmed access by the checkpoint means switch to reduced scope. |
| Owner and review | Anika checks access before commitment and reviews the result after the milestone. |
Do not invent numerical confidence just to fill a field. If you can give a defensible probability, record its basis. Otherwise describe the uncertainty and the observation that would change your view. Reserve the journal for choices whose reasoning is worth revisiting, not every routine approval.
Separate an unlucky result from a neglected condition
Suppose the contractor starts on time but an unrelated customer dependency delays the milestone. That result does not by itself show that choosing the contractor was wrong. Review whether the original evidence was adequate and whether the unrelated dependency should reasonably have been considered.
Now suppose access was never confirmed and Anika committed anyway. The delay exposes a process failure: the team ignored the recorded reversal condition. Calling that merely bad luck would miss a useful lesson.
A leader's review should ask both questions. What happened? And was the decision reasonable, including its execution, given what was known? Keep accountability for ignored evidence without pretending every unfavorable result proves negligence.
Preserve the entry, add the observed result, and identify one change to the next decision process. Do not rewrite the earlier note to make the later story cleaner.
Record one consequential choice before announcing it
Choose a decision with real uncertainty and a reviewable outcome. Write the alternative you did not choose and the assumption that would change your answer. Assign a review point now, not after the result becomes controversial.
At that review, compare the original reasoning with the evidence that later arrived. Look for a specific process adjustment rather than a verdict on whether you are a good decision-maker. For immediate release pressure, use a decision boundary before the outcome, not a retrospective journal alone.