Use one brief per test. This works for a flyer batch, a paid campaign, a customer reactivation effort or a commercial outreach experiment.
- Question: What are we trying to learn?
- Current evidence: What have we observed, and what is still an assumption?
- Audience and trigger: Who should see this, and why might they need the service now?
- Service and area: What is included, and what is excluded?
- Message and proof: What will we say, and what supports it?
- Destination and next step: Where will the person go, and what happens after contact?
- Costs: Media or distribution [amount]; external help [amount]; internal time [hours and valuation]; other costs [amount].
- Limit: Maximum cash exposure [amount], maximum effort [hours], and person responsible for checking both [name].
- Measurement: Source label, inquiry definition, fit criteria, booking definition and method for connecting outcomes.
- Timing: Launch date, technical check, early quality review and commercial review after the expected sales cycle.
- Decision rules: Conditions to continue, modify or stop. Include immediate stops for technical failure or a misleading promise.
Record the outcome in plain language
“We distributed [quantity] in [area], received [number] genuine inquiries, found [number] suitable and booked [number] jobs. [Number] remain pending. Total acquisition cost was [amount]. The main uncertainty is [issue]. Our next action is [decision].”
Do not omit zeroes or pending opportunities to make a test look better. If attribution is uncertain, say so. If several activities ran at once, avoid claiming all the work came from the one you prefer.
Keep the failed tests too. A useful record prevents you from paying to relearn the same lesson next season. It also makes handover easier if a colleague or provider takes responsibility later. They can see the reasoning, not just the current campaign settings.