Skip to content

Revenue leak

Diagnose where the funnel loses revenue.

  • Lesson 3 of 6
  • Intermediate
  • 6 min read
  • 2:16 video
  • Checked against the product 2026-11-03

Listen to this lesson · 2:15

In shortRevenue leak compares your GA4 shopping funnel (view item, add to cart, begin checkout, purchase) over the last 7 days with the 28 days before, finds the step where abandonment rose the most, and asks AI to explain it with device and browser detail and three fixes. Open Analytics → Google Analytics 4 → Revenue leak and click Run Diagnostics. It needs GA4 e-commerce events and costs credits per run.

How to use it in AI Rank Lab

  1. Find it in the dashboardAnalytics → Google Analytics 4 → Revenue leak
  2. Watch while you workOpen Revenue leak from here and the video plays in a small corner window of your dashboard, so you can follow each step on your own data. Expand it or go full screen any time. Open Revenue leak with the video
  3. Turn it into a routineThese playbooks show where this fits in a repeatable workflow:

What Revenue leak is for

Sales drop, and the first question is where. Did fewer people add to cart, or did more of them give up at checkout? Revenue leak answers that by comparing how many people drop out at each step of the funnel this week against your recent normal, so you look at the right page first.

Use it after a release, a theme or payment change, a price change, or any week when revenue looks off. It is a diagnosis tool, not a dashboard: run it when you have a question.

Before you start

  • GA4 connected and the shop's property selected. See Connect GA4 and read the overview.
  • GA4's recommended e-commerce events firing on your site: view_item, add_to_cart, begin_checkout and purchase. Shopify, WooCommerce and most store plugins can send these; check in GA4 under Reports → Engagement → Events.
  • About five weeks of data, so the 28-day baseline is full.
No e-commerce events, no result, credits still spent

If your property doesn't record those four events, every step comes back as 0% and the health badge can still say Healthy. The analysis is charged before the data is fetched, so confirm the events exist in GA4 before you run it.

Run it step by step

  1. Open Revenue leakGo to Analytics → Google Analytics 4 → Revenue leak. The page is titled Revenue Leak Diagnostician, with the price shown on a badge next to the button. Check the property in the toolbar is your shop.
    Revenue Leak Diagnostician page header with a credits badge and a Run Diagnostics button; below, a dashed info card describing the four funnel events and an empty state reading No analysis run yet.
    The empty Revenue leak page.
  2. Run the diagnosticsClick Run Diagnostics. The button changes to Analyzing... while GA4 data is fetched and the AI report is written.
  3. Read the three cardsWorst Funnel Step names the step whose abandonment rose most against the baseline. Current Checkout Abandon shows this week's drop-off from add to cart to begin checkout, next to the baseline. Despite the name, it is not the checkout-to-purchase step; read that one in the table below. Funnel Health is Critical when any step worsened by more than 20%, Warning above 10%, otherwise Healthy.
    Three result cards: Worst Funnel Step showing Add to Cart to Checkout with a plus percentage versus baseline, Current Checkout Abandon with a baseline figure, and a Funnel Health badge reading Warning.
    The summary cards after a run.
  4. Read the AI diagnosisThe AI Revenue Leak Diagnosis card leads with the biggest leak, points out device or browser patterns, lists three fixes in order of impact and gives an urgency level.
    AI Revenue Leak Diagnosis card with a short markdown report: a bold headline finding, a device and browser paragraph, three numbered fixes and an urgency line.
    The written diagnosis, generated from your funnel numbers.
  5. Check the tablesFunnel Step Analysis lists each step with current abandonment, baseline and the change. Event Counts shows the raw event numbers. Device & Browser Breakdown (Last 7 Days) shows where the events happened, so you can test the checkout on the device that looks worst.

How the numbers are worked out

NumberHow it's calculatedExample
Abandonment at a step1 minus (events at the next step ÷ events at this step)1,000 add_to_cart and 400 begin_checkout = 60% abandonment
BaselineThe same calculation for the 28 days before the last 750% abandonment in the baseline
VarianceHow much abandonment changed, relative to the baseline60% vs 50% = +20% variance

The current window runs from seven days ago to today, so it includes today's partial data. The baseline is the 28 days before that. Comparing rates rather than totals means the different window lengths don't matter.

Events, not people

The funnel counts events, not unique shoppers. Someone who views five products adds five view_item events. Small shops with a few dozen orders a week will see big swings from a handful of sessions, so treat a single Critical on low volume as a reason to look, not proof.

Confirm before you fix

Before changing your checkout, place a test order on the device and browser the report blames. Many "leaks" turn out to be a tag that stopped firing after a theme update, which looks exactly like shoppers leaving.

What it costs

Check yourself

1. Your GA4 property has no add_to_cart or purchase events. What will Revenue leak show?
2. Checkout abandonment was 50% in the baseline and 60% this week. What variance does the report show?
3. The report says Critical on a shop with 15 orders a week. What's the sensible next step?

0 of 3 answered.

Progress is saved in this browser.

Screenshots and video come from an AI Rank Lab demo account. Where a screen needs data that account doesn't have, such as a connected Google account or real visitors, it shows fictional example.com data.

Transcript
In this lesson you'll use Revenue leak to find where your shop's funnel is losing buyers. When sales drop, you need to know where. Fewer people adding to cart is a different problem from more people giving up at checkout. Revenue leak compares each step of your funnel this week with your recent normal, and points at the step that got worse. It needs GA4's standard e-commerce events: view item, add to cart, begin checkout and purchase. Most store platforms can send them. Check they exist in GA4 before you run this, because if they're missing you'll get zeros, and the credits are still spent. Go to Analytics, Google Analytics 4, Revenue leak. Check that the property in the toolbar is your shop. The price is on the badge next to the button. Click Run Diagnostics. After a short wait you get three cards. Worst funnel step names the step where more people dropped out than usual. Current checkout abandon shows how many people added to cart but didn't start checkout this week, next to the baseline. And Funnel health says Critical if any step got more than twenty percent worse, Warning above ten percent, or Healthy. Here's how the maths works. Abandonment at a step is the share of events that don't reach the next step. A thousand add to carts and four hundred checkouts means sixty percent abandonment. The tool compares the last seven days with the twenty-eight days before. If abandonment went from fifty to sixty percent, that's a twenty percent rise. Below the cards, an AI diagnosis explains the biggest leak, any device or browser pattern, three fixes and how urgent it is. Then there are tables with every step, the raw event counts, and a device and browser breakdown. Two cautions. The funnel counts events, not people, and small shops swing a lot from a few sessions. And a tag that stopped firing after a theme update looks exactly like shoppers leaving. So before you rebuild anything, place a test order on the device the report blames and check that GA4 still records each step. In the next lesson, you'll use the Anomaly engine to explain a sudden drop in traffic.

Put the lesson to work

Paste your site and get a free SEO, AEO and GEO audit with the fixes, then come back to the next lesson.