GA4 Funnel Exploration: How to Build a Conversion Funnel Analysis (Step by Step).

The conversion funnel is the most common question in web analytics: where are users dropping off? GA4's Funnel Exploration gives you a purpose-built answer — a step-by-step visualization of a multi-stage process with exact drop-off percentages between each step. But "Funnel Exploration" is easy to misread if you don't understand the open vs. closed distinction, and it is easy to build a funnel that looks plausible but quietly under-counts because the step events are configured wrong. This guide walks through every part of the setup.

What is Funnel Exploration?

Funnel Exploration is one of the five analysis types inside GA4's Explorations section. To reach it: open GA4, click Explore in the left nav, then select Funnel exploration from the template gallery. It lives entirely outside the standard reports — you won't find it in Acquisition, Engagement, or any of the built-in tabs.

Its purpose is narrower than a general Exploration: you define a sequence of steps (each tied to one or more events), and GA4 shows you how many users or sessions completed each step, how many dropped off, and the percentage conversion between adjacent steps. The output is a vertical bar chart where each bar represents one step — taller bars indicate more completions, and the gap between adjacent bars is your drop-off.

Funnel Exploration is the right tool when you already know the path you want to measure. If you want to discover what path users take without a hypothesis, Path Exploration is the better fit — it shows you the actual sequences users followed without you specifying them in advance. The distinction matters: funnel analysis answers "how many people completed checkout step 3?" while path analysis answers "what do people do after they view a product page?"

Open vs. Closed Funnels: The #1 Question

Every funnel you build in GA4 is either open or closed. This is the single most important configuration choice, and getting it wrong is the most common reason a funnel produces misleading numbers. The toggle sits at the top of the steps panel — it defaults to closed.

Closed funnels

In a closed funnel, users must enter the funnel at step 1. If a user completes step 2 without having completed step 1 first — within the same session or user journey — they are excluded entirely. Closed funnels measure the strict sequential path: did the user complete step 1, then step 2, then step 3, in that order?

This is the right choice for a checkout funnel where you genuinely want the step-1-to-completion rate. A user who lands directly on the order confirmation page without going through your cart is not a valid checkout conversion — they may have come from an email deep link, a cached URL, or a browser history shortcut. Counting them inflates your checkout conversion rate. Closed funnels exclude these users automatically.

The trade-off: closed funnels will count fewer users than open funnels on the same steps, which sometimes alarms teams who expect to see higher numbers. The lower count is more accurate, not a sign something is broken.

Open funnels

In an open funnel, users can enter at any step. A user who completes step 3 without having completed steps 1 or 2 is still counted at step 3. GA4 backfills their presence at the earlier steps they skipped, so the funnel shows them as having "entered" at the step they actually completed.

Open funnels are usually what you want for content or engagement funnels where strict sequence does not apply. A lead nurture funnel (read blog post → view pricing → contact sales) is genuinely open — a prospect who goes straight to pricing without reading the blog post is still a valid pricing visitor. Using a closed funnel here would artificially exclude a large portion of your audience and make the data look worse than reality.

For ecommerce, the guidance is less universal. A cart-to-checkout funnel benefits from being closed (strict sequence matters). A product discovery funnel (view category → view product → add to cart) is often better as open, because users regularly skip category browsing and land directly on product pages from ads or search.

The time window

For closed funnels, GA4 applies a 72-hour completion window by default. This means all steps must be completed within 72 hours of the user completing step 1. You can adjust this in the funnel settings — from as short as a few minutes (appropriate for a checkout flow a user should complete in one session) to as long as 90 days (appropriate for a long consideration cycle). For open funnels, there is no enforced time window — GA4 looks at the full analysis period.

How to Configure Funnel Steps

Each step in a Funnel Exploration is defined by one or more event conditions. The step is considered complete when GA4 observes the event — or the event plus the conditions you specify — firing for that user or session.

Simple event steps

The simplest step is just an event name: view_item, add_to_cart, begin_checkout, purchase. For a standard ecommerce funnel, these four events map directly to GA4's recommended ecommerce schema and should work out of the box if your implementation follows the spec.

Step naming matters. GA4 uses the step label — not the event name — as the axis label in the funnel chart. Rename "Step 1" to "Product page viewed," "Step 2" to "Added to cart," and so on. Anyone reading the funnel without deep knowledge of your event taxonomy can understand it immediately.

Event + parameter conditions

Where steps get powerful is when you combine an event with a parameter condition. This lets you filter a step to a specific page, product, or flow. To build a checkout funnel that only tracks the checkout page itself (not all page_view events), configure the step as: event = page_view AND page_location contains /checkout.

You can combine multiple conditions within a single step using AND logic. A step can require a specific event, a specific page path, and a specific item category simultaneously — only users who satisfy all conditions are counted at that step. This lets you build funnels that are specific to a product line, a campaign, or a user cohort without needing to create separate segments.

User vs. session counting

Funnel Exploration can count by users or by sessions. The choice affects your numbers significantly. User counting (the default) asks: of all users who hit step 1, how many eventually hit step 2? Session counting asks: of all sessions where step 1 occurred, how many sessions also included step 2? User counting is almost always the right choice for ecommerce funnels — you want to know whether a person bought, not whether the specific session they added to cart also contained a purchase (they may have returned the next day).

Trended vs. Standard View

The default Funnel Exploration view is the standard view — a snapshot of funnel performance across the entire selected date range. The step bars show total completions for the period, and the drop-off percentages are aggregated across all days.

Switch to trended view by clicking the chart type toggle at the top of the funnel canvas. Instead of a vertical bar per step, you now see a line per step plotted over time. This answers a different question: is this step's conversion rate stable, improving, or declining over time?

Trended view is particularly useful for A/B test analysis. If you ran a checkout page redesign and want to know whether it improved the add-to-cart-to-purchase conversion rate, plot both steps in trended view and look for a inflection point at the date the test variant launched. A clean lift in the step-3-to-step-4 conversion rate after the launch date — with no corresponding change in earlier steps — is good evidence that the redesign helped the specific step it targeted.

A third view option is elapsed time: it shows how long users take to move between each step. If your checkout funnel has a median elapsed time of 45 minutes between "begin checkout" and "purchase," that is a signal that users are abandoning and returning, which may mean the checkout flow has a friction point that sends users away to compare prices before returning to complete.

Using Segments and Breakdowns

A Funnel Exploration becomes diagnostic — not just descriptive — when you apply segments or breakdown dimensions to understand why users drop off at a particular step.

Segment comparison

Apply two segments to the same funnel to compare them side by side. The most common comparison is mobile vs. desktop: apply a "Device category = mobile" session segment and a "Device category = desktop" segment, and GA4 renders two sets of bars for the same funnel steps. If mobile users drop off at step 2 at a rate 40 percentage points higher than desktop, you have a specific hypothesis to test — the step-2 page likely has a mobile usability problem, not a messaging or pricing problem.

Other useful segment pairs: new users vs. returning users (are returning users more likely to complete checkout?), paid traffic vs. organic (does the paid cohort convert at a different rate?), or logged-in vs. logged-out users (does account authentication affect checkout completion?).

Breakdown dimensions

In the right-side configuration panel, you can apply a breakdown dimension to any step. A common use is adding device_category as a breakdown on the step with the highest drop-off — this shows the drop-off rate separately for desktop, mobile, and tablet within the same funnel. You can also break down by country, browser, or any custom dimension your implementation sends.

Next action after drop-off

At the bottom of the Funnel Exploration canvas, GA4 shows a "next action" panel for users who dropped off at a selected step. Click any step's abandon count, and GA4 shows what those users did next — what events fired, what pages they visited. If the majority of users who abandon at "begin checkout" go to page_view for your returns policy page, that is actionable insight: they want assurance before completing the purchase, and surfacing the returns policy on the checkout page itself may reduce drop-off.

Common Mistakes and Gotchas

Using the wrong event for a SPA

On single-page applications (SPAs), page_view does not fire automatically on navigation — only on the initial page load. If your checkout is a multi-step SPA flow where each "page" is a route change without a full browser reload, page_view events will only fire once. Steps defined as page_view + page_location contains /checkout/step-2 will show zero or near-zero completions. Use dedicated events (checkout_step_2_viewed, or the standard begin_checkout / add_payment_info / add_shipping_info ecommerce events) for SPA checkout steps.

Closed funnel under-counting from inconsistent step-1 events

For a closed funnel to count a user, GA4 must observe the step-1 event first. If your step-1 event fires inconsistently — only on some page loads, only after a JS race condition resolves, only when cookies are present — then many users who genuinely entered the funnel are excluded. The result looks like a very low top-of-funnel count with a deceptively high conversion rate. When a closed funnel shows a suspiciously strong step-1-to-purchase rate, the first thing to check is whether step 1 is firing reliably for all users, not just some.

Date range too short

For a closed funnel with the default 72-hour window, users who start the funnel near the end of your selected date range may not have had time to complete all steps before the window closes. A 1-day date range for a funnel where users typically take 24–48 hours to complete will show artificially low conversion at the final steps. Set the date range at least 2–3x longer than your typical completion time to give the funnel enough headroom.

Sampling

Funnel Exploration, like all Explorations, is subject to sampling on high-traffic properties or long date ranges. A sampled funnel may show conversion rates that differ from the true rate by several percentage points. Watch for the shield icon in the top-right of the canvas — if sampling is active, narrow the date range or reduce the breakdown dimensions before trusting the drop-off percentages for decisions. For more on how sampling works and how to work around it, see the guide on data sampling in GA4 Explorations.

Funnel Exploration vs. Path Exploration

Both tools live in the Explorations section, but they answer fundamentally different questions. Use Funnel Exploration when you already know the steps in advance and want to measure drop-off between them. Use Path Exploration when you want to discover what sequences users actually follow without imposing a hypothesis.

A practical decision rule: if you can name every step before opening GA4, use Funnel Exploration. If you are asking "I wonder what users do after they view the pricing page," open Path Exploration and let the data show you. Trying to reverse-engineer a discovery question into a Funnel Exploration — by guessing steps and seeing what shows up — is slower and less insightful than Path Exploration built for that purpose.

The two tools also complement each other. A common workflow: Path Exploration reveals that a large portion of users go from pricing to a specific feature comparison page before purchasing. You then build a closed Funnel Exploration with those three steps to measure the conversion rate precisely and track whether it changes over time.

Frequently Asked Questions

What is the difference between open and closed funnel in GA4?

In a closed funnel, users must complete step 1 before any subsequent steps are counted. In an open funnel, users can enter at any step — if they complete step 3 without completing steps 1 and 2, GA4 still counts them at step 3. Closed funnels are appropriate when the sequence is strictly required (checkout flows). Open funnels are appropriate when users may legitimately skip earlier steps (content or lead nurture funnels).

Can I use Funnel Exploration for ecommerce?

Yes, and it is one of the most common use cases. A standard ecommerce funnel uses view_item → add_to_cart → begin_checkout → purchase. Set it as closed if you want the strict cart-to-purchase conversion rate, or open if you want to include users who skipped directly to checkout from an ad landing page. The funnel requires that these events are implemented correctly in your GA4 property — if any step event is missing or fires inconsistently, the corresponding funnel step will be empty or under-count.

Why is my GA4 funnel showing zero users at step 2?

The most common causes are: (1) the step-2 event is not firing — check GA4 DebugView or the Realtime report to confirm the event fires when you manually complete that step; (2) for a closed funnel, the step-1 event is not firing reliably, so no users qualify to be measured at step 2; (3) the event parameter condition on step 2 does not match the actual parameter values your implementation sends — check the exact string including case, slashes, and encoding; (4) the date range is too recent for a closed funnel with a 72-hour window.

How many steps can a GA4 funnel have?

GA4 Funnel Exploration supports up to 10 steps. For most funnel analyses — checkout flows, lead capture sequences, onboarding flows — 3 to 5 steps is sufficient to identify where the largest drop-off occurs. Adding more steps beyond where meaningful drop-off happens dilutes the visual and makes the funnel harder to read without adding diagnostic value.