Read a sample: 1. Why Decisions Need a Memory
The project was supposed to launch at the end of the month. The team had finished most of the customer onboarding work, support staff were preparing their materials, and the sales team had begun talking to customers about the new service.
Then a supplier reported a delay in delivering a component the team needed for the launch. The revised delivery date was uncertain. The project leader had to choose whether to postpone the entire launch, release a limited version using a temporary workaround, or proceed on the original schedule and hope the component arrived in time.
None of those choices was comfortable. A delay would disappoint customers and keep staff tied to the project longer. A limited release could create extra work for the support team and exclude some customers. Proceeding as planned would protect the original schedule only if an uncertain delivery arrived in time.
The leader chose a limited release. The team would invite a small group of customers, use a manual process where the component was missing, and postpone the broader launch until the supplier’s delivery was confirmed. The decision seemed reasonable to the people in the meeting. They understood the pressures, knew which facts were uncertain, and agreed to check progress as the work continued.
Several weeks later, the supplier delivered earlier than expected. Someone asked why the company had not launched to everyone on the original date. Another person remembered that support had opposed the limited release. A third recalled that the manual process was supposed to last only a few days.
Each recollection contained a fragment of the discussion, but no one could easily reconstruct the choice as it had appeared when the team made it. The early delivery, now an established fact, made the original uncertainty harder to see.
This is the problem a decision record helps solve. It gives a decision a usable memory: a concise account of the choice, the alternatives, the information available at the time, and the reasoning that connected them. It also identifies who will act, what needs watching, and when the team should reconsider.
A record does not make uncertainty disappear. It makes the response to uncertainty easier to understand, communicate, and revise.
Leaders often encounter decisions twice. First, they make the choice with incomplete information and a deadline approaching. Later, they or others evaluate it with the benefit of events that have already unfolded.
Those are different vantage points. A sensible choice can lead to a disappointing result. A risky choice can work out because circumstances happen to favor it. The outcome matters, but it is not the only evidence of decision quality.
Imagine that the project team had proceeded with the full launch and the component had arrived in time. The launch might have gone well, but that result alone would not tell us whether the team had understood the delivery risk or prepared for a delay. If the supplier had missed the date, the same choice might have exposed customers and staff to a problem the team could have anticipated.
The limited launch has the same distinction. The supplier’s early delivery does not prove that the cautious choice was wrong. Nor does the existence of a written record prove it was right. The useful questions are more specific: What did the team know? How reliable was that information? Which consequences mattered most? What alternatives were available, and why did the leader choose one over the others?
A decision record preserves enough of that context to make those questions answerable. Think of it as a snapshot of reasoning rather than a verdict on performance. It is not a prediction, a guarantee, or a substitute for judgment. It is a practical aid for people who need to coordinate now and learn later.
The snapshot must be honest about its edges. It should not imply that the team considered every possible option or had access to information it did not have. If a choice was made quickly, the record can say so. If a critical point rested on an assumption, the record should name it. A short, candid account is more useful than a polished document that makes the decision appear more certain than it was.
A useful record answers a small set of questions:
- What was the decision? State the choice that had to be made, including any meaningful deadline or boundary.
- What options were considered? Capture the viable alternatives, including waiting or taking a limited step when those were real possibilities.
- What did the team know, and what was it assuming? Distinguish available evidence from the expectations used to fill gaps.
- What was chosen, and why? Connect the choice to the priorities and trade-offs that mattered at the time.
- Who owns the decision and the follow-through? Identify who can answer for the choice and who is responsible for important actions.
- When or under what conditions will it be revisited? Name a review point or an observable signal that could warrant another look.
These questions are a starting point, not a demand for an elaborate form. A decision involving several teams and a difficult reversal may need more explanation than a reversible choice within one team. The aim is to preserve what another person—or your future self—would need to understand the decision without reconstructing it from scattered messages.
For the disrupted project, an initial record might look like this:
>…