A prioritised answer to four questions is more useful than a 150-page audit. This is the structure of that answer, plus a worked hypothetical example.
Stop: activity that cannot be shown to create incremental value. Fix: things constraining everything downstream of them. Scale: things already working where more investment is defensible. Test: high-leverage bets where confidence is currently too low to commit.
Every item carries an owner, the metric it should move, an expected direction, and the evidence behind it. Items without a metric or an owner do not enter the plan.
An illustrative retailer with strong paid efficiency and weak repeat rate: stop broad prospecting expansion, fix post-purchase and lifecycle sequencing, scale the two profitable non-brand clusters, test a range-led category page structure. Figures are not used because the point is the sequencing logic, not invented results.
The view is revisited on a fixed cadence. Items move between columns as evidence changes; the column an item sits in is a statement about confidence, not a permanent judgement.
The worked example is hypothetical and illustrative. It is not a client engagement, and no client data or results are included.
Where this applies
This view is delivered at the end of every diagnostic.
View engagementWhere the plan needs ongoing senior ownership and re-prioritisation, embedded growth carries it.
View engagementMore from the library