Revenue Leak Radar: AI Analyst That Caught Broken Tracking on Day 1
đ¸ Real build, real screenshots, not mock-ups.
I built an AI analyst that checks a store's GA4 data every morning, spots revenue drops, funnel leaks and broken tracking, and explains each issue in plain English with a cost, an action and an owner.
The result: when revenue showed $0 but 19 customers had actually paid, the radar flagged it as a tracking problem on the first day, not a sales crash.
What it does:
⢠GA4 â BigQuery: 10 business metrics every day
⢠Compares each day to the same weekday over 28 days
⢠Catches orders recorded without revenue
⢠Gemini writes the diagnosis, action and owner
⢠Sends a visual report and an alert email
⢠"Ask your data" box: type a question, it writes and runs the SQL
Tools: Google Apps Script, BigQuery, GA4, Gemini API, Google Sheets
Cost: $0 (free tiers only)
Tested on Google's public GA4 demo dataset (Google Merchandise Store).
Full case study: https://claude.ai/artifact/KjLpN5RTwNLKTt3Ng77ssV

0 likes ¡ 7 views ¡ Oct 2, 2026
Revenue Leak Radar: AI Analyst That Caught Broken Tracking on Day 1 đ¸ Real build, real screenshots, not mock-ups. I built an AI analyst that checks a store...
Timeline
Sep 25, 2026 - Oct 2, 2026