Projects often remember the decision and forget the conditions that made it sensible. Months later, a new teammate asks why a feature was excluded or why a particular process exists. A small decision log preserves enough context to answer without rebuilding the conversation from memory.
Record decisions worth revisiting
Not every choice needs an entry. Capture decisions that affect scope, cost, ownership, delivery or the experience of the people using the work. If reversing the decision would surprise collaborators or require substantial rework, it probably deserves a short record.
Keep everyday execution choices close to the work itself. A decision log should not become a second task manager. Its purpose is to make the important reasoning discoverable, not to require formal justification for every adjustment to a sentence or meeting time.
Use a compact format
Start with the date, decision, owner and status. Add the situation, options considered and the main reason for the choice. Link to evidence when useful, rather than pasting entire reports. Include significant tradeoffs: what becomes easier and what the team accepts as a limitation.
A simple entry could say: “We will launch with three guides because the remaining topics require research. The editor approves publication. We accept a smaller initial catalogue and will review the next batch after source checks.” That record explains the boundary without pretending that the first release is the final destination.
Distinguish a decision from an assumption
An assumption can support a decision without becoming an established fact. Label it explicitly. “We assume most readers arrive on mobile” is different from “Our analytics show most readers arrive on mobile.” If no reliable measurement exists, leave the uncertainty visible and explain how it will be checked if it matters.
Likewise, distinguish recommendations from approved choices. A status such as proposed, accepted or superseded prevents an early suggestion from being mistaken for the current agreement. Use the vocabulary your team understands; consistency is more important than a formal template.
Add a review trigger
Some decisions should stand until new evidence appears. Others depend on a date, audience size, resource limit or unresolved test. Write the trigger that would justify revisiting the choice. This makes review purposeful and reduces the chance of reopening the same debate merely because someone was absent the first time.
A trigger might be “review when the second language is added” or “reconsider if the pilot reveals repeated access problems.” Avoid an arbitrary recurring review when nothing relevant is likely to change. A log should support work, not fill calendars with maintenance of the log itself.
Keep history without making it heavy
When a decision changes, create a new entry or mark the old one as superseded and link to the replacement. Preserve the original context rather than silently rewriting it. A decision that was reasonable under earlier constraints can look careless if those constraints disappear from the record.
Store the log where the project team already looks for information and link it from the main brief. Make ownership clear so that entries do not become a mixture of private recollections. For sensitive decisions, use appropriate access controls and avoid recording unnecessary personal information.
The log is useful when someone can answer three questions: what did we choose, why did we choose it, and what would make us reconsider? Pair it with a project handoff so the next person inherits the reasoning as well as the files. Begin with the next consequential decision, not an attempt to document the entire past.
Use it during a disagreement
When an old decision is challenged, read the entry together before defending it. Ask whether the original conditions still hold and whether new evidence changes the tradeoff. The log should make reconsideration fairer, not become a weapon for ending discussion with the claim that the matter was already decided.
If the challenge reveals missing context, add that context to the new record. If the decision remains appropriate, explain why and identify the agreed review trigger. People can then disagree with an explicit rationale instead of arguing against an imperfect recollection. A useful log supports accountable judgement while allowing the work to evolve as circumstances change.