Marketing Campaign Diagnostic Engine
Systematically diagnose underperforming marketing campaigns by separating surface symptoms from root causes using empirical data.
Customise & copy
Add the context you have. The prompt updates as you work.
State the primary goal and key performance benchmarks.
Provide spend, impressions, CTR, CPC, conversion rate, CAC, and ROAS.
Describe who the campaign was targeted at.
Paste the ad copy, headline, or describe the visual creatives.
Select the primary advertising channel used.
Live prompt
5 fieldsYou are an expert growth marketing diagnostician and data analyst. Your task is to diagnose an underperforming marketing campaign objectively and systematically, separating surface-level symptoms from root causes without inventing missing data. Campaign Objective and Target KPIs: Generate B2B SaaS demo requests at a target CAC under $150 and ROAS of 3.0. Quantitative Campaign Metrics: Spend: $10,000, Impressions: 500,000, CTR: 0.5%, CPC: $4.00, Conversion Rate: 1.0%, CAC: $400, ROAS: 0.8 Audience Definition and Targeting Parameters: Job Title: VP of Marketing, Industry: Enterprise Software, Location: US, Lookalike Audience (1%). Creative Assets and Messaging Copy: Ad copy: 'Scale your marketing pipeline instantly with our AI platform.' Static image featuring software dashboard screenshot. Channel-Specific Performance Data: LinkedIn Ads Analyze the provided data strictly and output your diagnostic report in the following format: 1. FUNNEL DROP-OFF ANALYSIS - Identify the exact drop-off point in the conversion funnel using the provided metrics. - Separate surface-level symptoms (e.g., low CTR) from underlying systemic causes (e.g., poor audience-market fit). 2. ROOT CAUSE HYPOTHESES - List the primary root causes strictly supported by the empirical data provided. - Avoid jumping to causal conclusions based purely on correlation. 3. PRIORITIZED HYPOTHESES - Rank the hypotheses strictly by the strength of the supporting evidence. 4. VALIDATION TESTS - Provide a minimal, high-leverage set of A/B or diagnostic tests to validate the leading hypotheses.
Proof before you use it
A real example
Tested on gemini-3.1-flash-lite on 2026-09-06
{
"channel_data": "TikTok Ads",
"creative_assets": "Fast-paced TikTok video meme format addressing student loan debt. Copy: 'Stop losing money to hidden fees.'",
"campaign_metrics": "Spend: $15,000, Impressions: 3,000,000, CTR: 0.8%, CPC: $0.62, Conversion Rate: 15% (Install rate), CPI: $4.15",
"campaign_objective": "Acquire mobile app installs for a fintech budgeting app with a target cost-per-install (CPI) under $3.00.",
"audience_definition": "Gen Z and Millennial students and young professionals interested in personal finance and saving money."
}A small ritual that works
How to use it
- 01Input your campaign's objective, spend, and conversion metrics.
- 02Paste your current ad copy and audience targeting parameters.
- 03Run the prompt to receive a structured funnel drop-off analysis.
- 04Execute the recommended high-leverage validation tests to optimize performance.
Why it works
The prompt enforces a rigorous, structured diagnostic process that separates surface-level marketing symptoms from underlying funnel bottlenecks. By systematically requiring a funnel drop-off analysis, empirical root cause hypotheses, evidence-based prioritization, and minimal high-leverage validation tests, it prevents marketers from jumping to speculative conclusions based on isolated metrics.
Where it fails
INSUFFICIENT_EVIDENCE
Customisation tips
Ensure the input metrics provided in `campaign_metrics` are mathematically consistent and internally reconciling (e.g., matching impressions, CTR, CPC, and spend) to prevent analysts from building diagnostics on flawed baseline calculations. Extend the diagnostic framework to include secondary conversion events or multi-channel attribution data if richer tracking is available.
Watch-outs
- Do not invent external industry benchmarks (e.g., standard CVR or CTR for specific verticals) that are not present in the provided campaign data.
- Avoid inferring landing page friction, pricing mismatches, or audience quality without explicit data on user behavior post-click.
- Never present speculative hypotheses as definitive root causes; strictly rank them by the empirical strength of the provided metrics.
Works in
Was this useful?
Submit your variation
If your version helps, it may be published with your name and a link back to you.
Want to make AI useful across your team? Explore Atul's training.