Post-campaign reporting, without the manual deck
Problem
Every campaign wrap meant a day or more of pulling numbers from three platforms into a deck by hand — slow enough that decisions on what to run next often happened before the report did.
What I built
A reporting pipeline that pulls platform + attribution data on a schedule, applies decision-rule logic (kill/scale/hold thresholds by channel and funnel stage), and outputs a ready-to-present deck instead of a raw export.
Outcome
Post-campaign turnaround went from multi-day to same-day, and the kill/scale call became a rule applied consistently instead of a judgment call redone from scratch each time.