Idea work is stronger when generation and evaluation use clear criteria. Brainstorming produces many options, but without criteria the loudest idea wins.
Generate broadly, then score by value, effort, risk, differentiation and testability.
Start with a narrow boundary: which website, space, file, recipient or decision is affected? This makes the task reviewable instead of turning it into a broad catch-all request.
A useful work order is: “Collect ideas for this goal and evaluate them by value, effort, risk, differentiation, testability and next experiment.” For important cases, add that uncertainties must be marked visibly instead of being filled in silently.
Pay special attention to current work status, confirmed decisions and open questions. These points decide whether the result is only useful for the moment or can be found, checked and continued by the team later.
Do not treat the first creative list as a plan.
The team gets both creative range and a practical shortlist.