Begin with constraints and leaks

Check whether stock, fulfillment, margin, or measurement problems would limit the value of a new campaign. Sending more demand to a product about to run out can waste spend; scaling a discounted product with weak contribution can make the topline look healthier while cash deteriorates.

Fix urgent revenue leaks and capacity constraints before adding another growth experiment.

Turn observations into actions

A useful priority is specific enough to complete. Replace 'improve inventory' with 'reduce promotion of the two products with less than 14 days of cover' or 'build a bundle around the slow-moving accessory with strong attach potential.'

Attach the evidence, expected result, owner, and review date. This makes a recommendation testable instead of merely interesting.

Rank a small queue

Score expected impact, confidence in the evidence, effort, urgency, and reversibility. A fast, reversible merchandising change supported by current order data may deserve priority over a larger but speculative project.

Limit the active queue. Completing two meaningful actions creates more learning than carrying twelve priorities into another week.

Close the weekly loop

  • Record the baseline before making the change.
  • Assign one owner and one review date.
  • Separate leading signals from final commercial outcomes.
  • Keep, revise, or stop the action based on evidence at review time.