Meetings
Part of Decision efficiency
Preventing repeated debate over a settled decision
A team may revisit a decision because the result was never recorded, because implementation is hard or because genuinely new facts emerged.
A team may reopen a decision because the result was never recorded, because implementation is hard or because genuinely new facts emerged. Treat those reasons differently: a decision log should allow correction without forcing everyone to relitigate the same arguments each week.
Preserve the original reasoning
Record the question, the chosen option, the alternatives considered, the key evidence, the decision-maker and the effective date. State the condition for review, such as a measured failure or a changed legal requirement. Atlassian's DACI framework emphasises sharing the decision once roles and input are resolved, and the shared record is what makes that possible.
When someone proposes reopening the decision, ask what has changed. A new customer constraint or a failed test may justify review. A concern already discussed should be acknowledged, but the team can point to the recorded trade-off and continue implementation. If the original decision had no evidence or authority, correct the process rather than pretending it is settled.
When to Reopen a Decision vs. When to Continue Execution
- Reason to reopenNew facts (e.g., failed test, new customer constraint, changed law)
- Reason NOT to reopenConcerns already discussed and documented in the decision log
- Implementation difficultyNot a valid reason to revisit a decision—address via action lists and support
Give execution its own forum
Many ‘should we still do this?’ conversations are really requests for help with implementation. Move them to an action list with owners, dates and obstacles. Front-line staff should not have to defend a strategic decision they did not make; route material objections to the person with authority to revisit it.
Set a planned review for choices that are uncertain and reversible. At that point, compare actual results with the criteria written before launch. This gives people a legitimate way to learn and change course without constant second-guessing.
