Funnel Teardown: Finding the Real Drop-Off in a GA4 Checkout Funnel
Objective: Given a 4-step GA4 Funnel Exploration output, identify the real drop-off point and rule out plausible-sounding but wrong explanations.
You're reviewing a GA4 Funnel Exploration report for Care.com's premium membership checkout, built from Product View to Purchase.
Read the funnel percentages, find where the real leak is, and reject the distractor explanations that don't match the data.
Before you start
What you'll need
Free path (everything below is enough to finish)
Explore > Funnel exploration is included in every free GA4 property
Connects directly to GA4 at no cost
The process
Specimens to review
Where is the real leak in this funnel, and what in the data rules out the most obvious wrong answer?
GA4 FUNNEL EXPLORATION: Premium Membership Checkout Step 1 Plan Page View 10,000 users 100% Step 2 Add to Cart 6,400 users 64.0% (-36.0%) Step 3 Begin Checkout 5,760 users 90.0% of step 2 (-10.0%) Step 4 Purchase Complete 1,210 users 21.0% of step 3 (-79.0%) Elapsed time, Step 3 to Step 4: median 6 minutes 40 seconds Device split at Step 4: 78% mobile, 22% desktop
Specimen: synthetic, realistic
Final deliverable
A one-paragraph teardown naming the real leak, the ruled-out distractor, and the supporting evidence (elapsed time, device split).
See a reference example
Funnel Teardown, Stitch Fix Style Pass Checkout The real leak is the 74% drop between Begin Checkout and Purchase Complete, not the smaller drop at Add to Cart. Median time-on-step of over 5 minutes at that stage points to active friction, likely form length or a surprise fee, rather than a browsing decision. The Add to Cart drop is a distractor: it has the biggest raw number but sits within a normal range for a first-touch action. Fix the checkout step first.
Success criteria
You're done when you can:
- Names the checkout-to-purchase step as the real leak, not the largest raw-number drop
- Uses the elapsed-time and device-split evidence to support the diagnosis, not just the percentages