Ask five people in an ecommerce business what the priority is for the next quarter and you will usually get five defensible answers. That is not a communication failure; it is what happens when no plan has been made explicit enough to disagree with.
One named constraint
A single sentence, agreed by the commercial owner: the thing limiting growth this quarter. Not three themes. If it cannot be stated without an 'and', the diagnosis is not finished.
Decisions, not activities
Each item should be a decision with an owner, a date and a consequence — repricing a range, retiring a channel, rebuilding a template, changing a brief. 'Improve PDPs' is an activity. 'Ship the new PDP template for the top 40 SKUs by week six, measured on add-to-basket rate' is a decision.
An explicit stop list
- What is being paused for 90 days, and who has been told.
- Which tests will not run because they would not change a budget decision.
- Which reporting is being retired so the remaining numbers get read.
- Which agency scope shrinks or sharpens as a result.
A plan without a stop list is a wish list with dates on it.
One measure per decision
Each decision carries one number that would tell you within the quarter whether it worked. Two numbers means an argument; none means it survives on opinion.
Written this way, a 90-day plan is short — usually a page. It is also the artefact that makes existing agencies noticeably better, because they finally receive a brief that reflects a commercial choice.