Deep Dive Into GA4 Funnel Drop-off Analysis

GA4's Funnel Exploration is one of the most useful reports the platform offers, and one of the most underused. Most teams look at it briefly, notice that there's a 60% drop-off at step two, and then don't dig much further.

The drop-off percentage is just the beginning. The useful analysis is understanding who's dropping off, when, and whether those drop-offs represent fixable problems or expected behavior.

Setting up a Funnel Exploration Correctly

Go to Explore in GA4, create a new exploration, and select Funnel Exploration from the technique options.

Your funnel steps should map to real events that your site fires when users take meaningful actions. On an e-commerce site, a typical conversion funnel might look like:

1. Product page view
2. Add to cart
3. Begin checkout
4. Purchase

On a SaaS site:

1. Landing page visit
2. Pricing page view
3. Trial signup
4. First meaningful action in-product

The events you use need to actually exist in your GA4 data. Before building the funnel, confirm in the DebugView or Events report that each event is being fired reliably.

The Open vs. Closed Funnel Distinction

GA4 lets you build either an open funnel or a closed funnel.

An open funnel includes users who enter at any step — not just users who started at step one. This is useful when users can legitimately enter your funnel at any point (e.g., some users navigate directly to the checkout page).

A closed funnel only includes users who started at step one and progressed through the steps in order within the session. This gives you a stricter measure of the sequential conversion experience you're designing for.

For most conversion optimization work, you want a closed funnel so you're measuring the experience of users going through the flow you designed, not all users who happened to touch those pages in any order.

Reading the Drop-off Numbers Correctly

Each step in the funnel shows you what percentage of users from the previous step made it to this step. An 85% drop-off from "add to cart" to "begin checkout" means that 15 in every 100 people who added something to their cart proceeded to checkout.

Before diagnosing this as a problem, ask whether it's actually lower than expected. Some cart behavior is browse behavior — people add to cart as a way of saving, or to see the total with shipping, with no strong intent to purchase in that session.

Compare your rate against industry benchmarks, but more importantly, compare it against your own historical performance. Is this rate worse than it was six months ago? If so, something changed. If it's been consistent for a year, it might be normal behavior for your audience and product.

The Breakdowns That Reveal the Real Story

The powerful part of Funnel Exploration is the ability to break down each step by a dimension.

Device category. Split by mobile vs. desktop vs. tablet. Many funnels look radically different across devices. If your mobile checkout completion rate is 40% lower than desktop, that's an actionable problem — mobile checkout has friction that desktop doesn't.

Traffic source. Break down by session source/medium or by first user source. Paid search traffic might convert at a very different rate through the funnel than organic traffic or email traffic. If email subscribers complete the funnel at twice the rate of cold paid traffic, that affects how you think about customer acquisition costs.

New vs. returning users. First-time visitors often have lower funnel completion rates because they haven't established trust with the brand. Returning users who've visited before converting show you the natural sales cycle length.

Geography. If you're selling internationally, conversion rates often vary significantly by country due to payment method availability, shipping costs, currency trust, or even the time zone (checkout abandonment tends to spike late at night).

Session-scoped vs. Event-scoped Funnels

GA4 allows you to configure funnel steps as either session-scoped (must happen in the same session) or event-scoped (can happen across multiple sessions within a date range).

For immediate conversion flows — someone visits and buys in one session — session-scoped is appropriate. For longer evaluation cycles where users visit multiple times before converting, event-scoped captures the full conversion path more accurately.

B2B products and high-consideration purchases typically need event-scoped funnels. An event-scoped funnel that spans 30 days will show you that buyers who visit five times before converting exist, and how many of them complete the purchase.

What to Do After Identifying a Drop-off

Finding the step with the highest drop-off is step one. Step two is forming a hypothesis about why it's happening.

A drop-off from product page to add-to-cart might mean: the product page isn't building enough confidence (missing information, no reviews, unclear sizing), the price is above what visitors expect, or the wrong audience is reaching the product page.

A drop-off from add-to-cart to checkout might mean: shipping costs visible at checkout are surprising, the checkout page loads slowly, or the guest checkout experience is buried.

A drop-off at payment entry might mean: a form usability problem, a payment method gap, or a trust signal missing at the moment of card entry.

Each hypothesis leads to a different investigation. Use session recordings (from Hotjar, Microsoft Clarity, or a similar tool) to watch actual users go through the steps you're investigating. The recordings often reveal friction that isn't visible in the numbers.

Connecting Funnel Data to Revenue

GA4's funnel exploration shows you percentages and user counts. To understand the revenue implications, you need to connect those numbers to your average order value and your current traffic.

A rough calculation: if fixing the drop-off from add-to-cart to checkout improved that step by 10%, and your current monthly revenue from that funnel is $50,000, the revenue opportunity is roughly $5,000 per month — assuming everything else stays constant, which it won't perfectly, but gives you a sense of the stakes.

This kind of rough revenue opportunity sizing is useful for prioritizing which drop-off to investigate first. Not all steps in the funnel have equal revenue potential.