marketing-performance-report

Provide your exported campaign, channel, and budget files, then specify the reporting period, timezone, currency, and targets. The Skill builds a metric dictionary, validates data quality, and produces a scorecard with funnel, budget, and attribution analysis plus prioritized decisions.

Compare only compatible periods with matching definitions, timezones, and currencies. Partial months must be compared against compatible partial periods or flagged as incomplete, and KPIs with different formulas or attribution windows should never be directly compared.

No. The Skill only analyzes data and produces reports; it never adjusts budgets, pauses or scales campaigns, updates lifecycle statuses, or sends reports. Saving a report file requires explicit user approval and runtime support.

Differences usually come from mismatched formulas, sources, attribution windows, or timezones. The Skill resolves this by building a metric dictionary and reconciliation table that lists differences, likely causes, and which version is adopted.

Attribution models like first-touch, last-touch, or position-based show association, not causation. Influenced pipeline cannot be summed as caused revenue, and without an experiment or holdout design, results are described as associated rather than driven.

Avoid it when you only have vanity metrics but want to claim ROI, when no baseline exists to show movement, or when you need statistical experiment inference, which belongs to experiment design instead.