Analyze ad performance, wasted spend and budget allocation across channels

1. Provide campaign data as a CSV export from Google Ads, Meta Ads Manager or LinkedIn Campaign Manager, a pasted table, or a dashboard screenshot. 2. Include the period, monthly budget, primary goal (demos, trials, purchases or leads), and any target CPA or ROAS. Example: “Here are my March Google and Meta campaigns with a $10,000 budget. I want demo bookings below $150. Where is spend wasted, and how should I reallocate it?” 3. If available, add funnel data to calculate adjusted acquisition cost: lead-to-MQL, MQL-to-SQL and SQL-to-close conversion rates, plus average deal value. Mention changes during the period and spending constraints. 4. The agent normalizes the data into one table, checks campaign health, finds waste such as spending without conversions or acquisition costs exceeding three times the target, identifies winners, and checks significance using minimums of 100 clicks per variant for CTR tests and 30 conversions per variant for CPA tests. 5. It produces a Markdown report with an executive summary, waste analysis, conservative and aggressive budget scenarios, and actions for the current week, current month and next month. It saves campaign-analysis-[YYYY-MM-DD].md in the working directory or a path you specify.

No external tools, network connection or API keys are required; analysis uses only the data you provide. The skill does not connect directly to ad accounts, so you must export the data. Screenshot reading requires an image-capable agent; saving the report requires write permission. Benchmarks and future estimates, including marginal returns and scenario outcomes, are reasoned estimates, not guaranteed results. Accuracy depends on completeness and sample size. This skill is not intended for planning new campaigns or generating ads without actual performance data.