When performance falls, every department suggests a solution from its own angle: advertising wants new creative, sales wants more leads, and operations wants less pressure. Diagnosis puts them on one chain and finds the largest drop before prescribing a fix.
This framework is distilled from the published case studies. It guarantees no numerical outcome; it reduces guessing and defines a measurable test.
Days 1–2: map the full journey
Start at the impression and end at delivery and profit. Collect impressions, clicks, conversations, orders, confirmations, deliveries, revenue and margin.
Do not search for the prettiest number. Find the largest gap between two stages; that is usually the first constraint.
- Offer
- Message
- Advertising
- Sales
- Operations
Days 3–4: listen to objections and operations
Read customer conversations, cancellation reasons, delivery notes and stock-outs. Quantitative data says where the drop happened; language helps explain why.
Many messages are not automatically success. An unclear message may force every customer to ask basic questions.
Day 5: write a falsifiable hypothesis
Replace ‘the campaign is weak’ with: if the difference is clear in the first five seconds, order progression will improve with audience held constant. The hypothesis identifies variable and metric.
Choose the hypothesis closest to the largest leak and cheapest to test.
Days 6–7: design the test and decision rule
Change one variable, define sample and duration, and record before and after. Agree in advance what outcome means continue or stop.
Scale only when the final profit-linked metric improves—not an intermediate number alone.
Conclusion
Good diagnosis does not hand you a long idea list. It gives one clear decision: where to begin, what to test and how to know it worked.