付费投放实验与扩展决策设计
围绕假设、渠道经济性和归因边界规划投放实验,输出测试矩阵、护栏指标与扩展规则。
完整提示词
Main objective: > Perform Campaign Experiment Design through a clear hypothesis, economics, experimental design, attribution limits and scaling decisions. ## 1. PAID CONTEXT Define objective, audience, offer, channel, budget, conversion event, attribution window, unit economics, creative inventory and landing experience. ## 2. PAID STANDARD - optimize to the deepest reliable business event available - separate platform attribution from incrementality - define test hypothesis before spend - isolate major variables where practical - include creative, audience and landing-page interaction - scale only after economics and operational capacity support it - use holdouts/geo/tests where appropriate for incrementality ## 3. EXPERIMENT CARD ```text Hypothesis: Channel: Audience: Creative: Landing page: Budget: Primary metric: Guardrail: Attribution assumption: Incrementality test: Scale/stop rule: ``` ## 4. REQUIRED MATRICES ### Paid Channel Matrix | Channel | Audience fit | CAC/CPA | Quality | Scale | Attribution risk | |---|---|---|---|---|---| ### Experiment Matrix | Test | Hypothesis | Variable | Primary metric | Guardrail | Decision rule | |---|---|---|---|---|---| ## 5. FAILURE MODES Avoid optimizing to cheap but low-quality conversions, changing many variables at once, reading platform attribution as causality, scaling before product/sales capacity and declaring winners from noisy short tests. ## 6. REQUIRED OUTPUT 1. Objective/economics. 2. Channel/test strategy. 3. Creative/landing hypotheses. 4. Attribution assumptions. 5. Required matrices. 6. Scale/stop rules. 7. Risk/guardrails. 8. Learning agenda. End with **Paid Acquisition Integrity Check** confirming decisions are tied to economics and test evidence, not platform vanity metrics. This prompt does not guarantee ROAS, CAC or incremental growth.
分类与标签
商业营销、付费投放、广告实验、归因分析、获客增长