Funnel Visualization
Funnel visualization is how you turn raw step-by-step conversion data into a chart that reveals where users drop off. The chart type you pick determines which optimization opportunities you actually see.
Funnel Visualization
Charting step-by-step conversion data — typically as step funnels, Sankey diagrams, or retention curves — to expose drop-off and prioritize fixes.
Funnel visualization is the practice of rendering sequential conversion data as a chart so drop-off between steps becomes visible at a glance. The three workhorse formats are the classic step-funnel bar chart (one bar per step, descending), the Sankey diagram (flows branching between paths), and the retention curve (users remaining over time or sessions).
Each format encodes a different question. A step funnel answers "where do most users leave?" A Sankey answers "which path did they take instead?" A retention curve answers "how quickly do they leave?" Picking the wrong chart hides the very leak you're trying to fix, which is why funnel visualization sits at the centre of any serious funnel analytics workflow.
On a Shopify storefront the canonical funnel is product view → add to cart → checkout start → purchase. A step-funnel chart shows each stage as a shrinking bar, with the conversion rate between adjacent steps labelled on top. It's the fastest read for "which single step leaks the most" — and it's usually the right starting point in funnel optimization.
Step funnels break down when users don't move in a strict sequence. If a beauty shopper bounces between PDP, reviews, quiz, and cart three times before buying, a linear funnel collapses that into one row and hides the real behaviour. That's where a Sankey diagram earns its place: it shows the actual flows, including the loops and the abandonment exits.
step_conversion_rate = users_completing_step / users_entering_step
users_entering_step
Step entries
Unique users who reached the start of this funnel step in the time window.
users_completing_step
Step completions
Unique users from that same cohort who completed the step's success event.
step_conversion_rate
Step conversion rate
The proportion of entrants who advanced — the number rendered above each transition in a step-funnel chart.
An apparel Shopify store measuring add-to-cart → checkout-start over a 7-day window.
Users entering step (added to cart): 4,200
Users completing step (started checkout): 1,470
→ 35%
35% cart-to-checkout is mid-pack for apparel; below 30% usually signals a shipping-cost or account-required friction worth testing first.
Read the chart in two passes. First scan the absolute drop between adjacent steps to find the biggest leak. Then compare that leak's conversion rate against benchmarks for the same vertical and platform — a 35% cart-to-checkout that feels bad is actually normal for apparel, while a 70% checkout-to-purchase that feels fine is poor for the same category.
Funnel visualization formats — when each one earns its place
| Chart type | Best for | Typical use case | Limitation |
|---|---|---|---|
| Step-funnel bar chart | Linear, well-defined funnels | Shopify checkout flow, signup wizard | Hides non-linear paths and loops |
| Sankey diagram | Branching journeys, multi-path flows | Mapping PDP → cart vs PDP → quiz → cart | Gets cluttered above ~8 nodes |
| Retention curve | Time- or session-based decay | Post-purchase repeat-buy rate over 90 days | Doesn't show which step caused churn |
| Cohort funnel grid | Comparing segments side by side | Mobile vs desktop checkout completion | Requires enough volume per cohort |
| Heatmap-on-funnel | Tying drop-off to on-page behaviour | Cart page exit + scroll depth overlay | Needs both quant and qual data joined |
The most common mistake is treating the chart as the answer. A funnel visualization tells you where users leave, not why — that requires session replay, on-page surveys, or a hypothesis test. Pair the chart with one qualitative signal before you commit a sprint to the fix, and segment by device, traffic source, and new-vs-returning before declaring a winner.
Funnel visualization FAQ
A funnel chart assumes a strict linear sequence and shows one bar per step. A Sankey diagram shows actual flows between any two events, including branches and loops. Use a funnel chart for checkout; use a Sankey when you want to see which path users took to get there.
Three to six steps is the readable range. Fewer than three and you're just plotting one conversion rate; more than six and the bars compress until you can't see the meaningful drop-offs. If your real funnel is longer, split it into two charts at a natural seam (acquisition vs checkout).
Show both. Absolute counts tell you whether a leak is worth fixing (10% drop on 100k users matters more than 40% drop on 500). Step-conversion percentages tell you how the funnel compares to benchmark. Most modern funnel analytics tools layer both onto the same chart.
Yes, GA4 has Funnel Exploration under Explore, which renders a step-funnel chart with optional segment overlays. It works for up to 10 steps and supports open vs closed funnels. The limitation is segmentation depth and the inability to render Sankey-style branching without exporting to BigQuery.
A closed funnel only counts users who entered at step 1; a user joining at step 3 is invisible. An open funnel counts users at whichever step they entered. Closed funnels give a cleaner conversion read; open funnels reflect reality when users land mid-flow from email or paid ads.
Use a Sankey diagram or a flow chart that allows branches. Define the events (not steps) you care about, and let the chart render the actual transitions between them. This is the right format for browsing-heavy journeys like beauty or apparel where users loop between PDP, reviews, and cart.
Usually because of attribution windows, deduplication rules, or whether the funnel is open or closed. A funnel exploration counts users who completed steps in sequence within a session-or-conversion window; a standard report counts the event regardless of order. Align the definitions before debugging the data.
When the question is about time, not sequence. Repeat-purchase rate, subscription renewal, or feature re-use over weeks all read better as a retention curve. Funnels handle one user journey; retention curves handle the decay of a cohort over time.
Weekly for active optimization, monthly for executive reporting. Daily refreshes invite over-reaction to noise — small stores rarely have enough daily volume for step-conversion rates to stabilize. Pair the visualization with a rolling 7- or 28-day window to smooth weekend effects.
It points to the leakiest step, which is where testing has the highest expected lift. It can't tell you the cause — for that you need session replay or a survey on the abandoning step. Modern CRO platforms increasingly pair the chart with AI-generated hypotheses based on the observed drop-off pattern.
Get an AI expert review of your site
Paste your URL — Metricuno's AI runs the same heuristic checks a senior CRO consultant would, scoring your page and prioritising the fixes that'll move conversion fastest.