Funnel Visualization: How to Build and Read a Funnel Chart

By Christoph Olivier, Founder, CO Consulting

Last reviewed: July 2026

Funnel visualization means turning your stage counts into a chart that shows how many people survive each step and where the biggest drops happen. This guide is about the chart itself: how to structure the data, which tool to build it in, and how to read the drop-off once it renders. It does not re-define the funnel stages or teach you to run the funnel week to week. It teaches you to draw the picture and read it correctly.

What funnel visualization actually shows

Funnel visualization shows the volume surviving at each sequential stage as a series of bands that get narrower top to bottom. The top band is the widest because it holds every entrant. Each lower band is thinner because fewer people reached it. The width difference between two bands is the drop-off, and the point where the width collapses hardest is your worst leak. That is the entire job of the chart: make the biggest leak visible without math.

A funnel chart is not a bar chart lying on its side. A bar chart compares independent categories. A funnel chart shows a dependent sequence where each stage can only draw from the one above it, so counts can only stay flat or fall. If a lower band is ever wider than the one above it, your data is wrong or your steps are out of order.

Once you can read where the funnel pinches, the fix work lives elsewhere. Our conversion funnel optimization audit covers diagnosing and repairing the leak the chart exposes.

Structure the data before you touch a chart tool

Funnel data is two columns: one for the ordered stage names, one for the count of people or events at each stage. List stages in strict sequence, top of funnel first. Add a third column for conversion percentages so the chart carries proportion, not just raw volume. Get these three columns right and every tool below will render a clean funnel in one step.

Include two percentages, because they answer different questions. Stage-to-stage conversion is (next stage count divided by current stage count). Overall conversion is (final stage count divided by first stage count). The first tells you which single step leaks. The second tells you the health of the whole path.

Keep the funnel to four to seven stages. Fewer than four and the chart says nothing a sentence could not. More than seven and the drop between adjacent bands gets so thin you cannot act on any single step. If you need the full stage list first, see our map of the real marketing funnel stages, then collapse it to the handful that matter for your view.

How to build a funnel chart in each common tool

The tool depends on where your data already lives. If numbers sit in GA4, build in GA4. If they span ads, CRM and analytics, build in Looker Studio. If you exported to a spreadsheet, Excel or Sheets renders a funnel natively. All four take the same two-column structure and turn it into the same narrowing shape.

1. GA4 Funnel exploration. Open Explore, choose the Funnel exploration template, and define each step as an event or condition. GA4 returns a stepped bar chart plus a table with per-step user counts and drop-off rates. Use this when your stages are on-site events like page views, sign-ups and purchases. Set it up correctly with our practical GA4 setup walkthrough.
2. Looker Studio funnel chart. Google added a native funnel chart in September 2024 with three styles: smoothed bar, stepped bar and inverted triangle. Pick a stage dimension and a metric, and it blends GA4, ad platforms and CRM into one funnel. Use this when the funnel crosses several sources.
3. Excel. Enter stage and value columns, select them, then Insert, then Waterfall, Funnel, Stock, Surface or Radar, then Funnel. Excel draws a centered horizontal funnel with no formula work. Use this for a fast static funnel from an export.
4. Google Sheets. Sheets has no built-in funnel type, so use a stacked bar chart with a transparent spacer series to center the bars, or a template. Use this when the team already lives in Sheets and you want a shareable link.

For a slide or a report cover where the shape matters more than live data, a design tool like Canva or Lucidchart gives you a clean inverted pyramid you type numbers into. That is a picture, not a data connection, so it will not update. Do not use it as your working funnel.

How to read a funnel chart without fooling yourself

Read a funnel chart by comparing absolute drop, not just the percentage. A step that loses 30% of 10,000 people sheds 3,000 and outranks a step that loses 50% of 500 and sheds 250. The percentage flags the leakiest step; the raw count tells you which leak is worth your week. Fixing the bigger absolute drop returns more revenue even when its percentage looks milder.

Then separate a leak from a design floor. Some steps are supposed to shed most people, and a low conversion there is normal, not broken. A visitor-to-lead step that converts at 2 to 4% is healthy for cold traffic. A booked-call to proposal step that converts at 60% is expected. Compare each step to its own benchmark, not to the step above it. Our conversion rate benchmarks give you the reference numbers per stage.

Last, remember the chart ends the analysis; it does not do it. A funnel chart tells you which step leaks. It never tells you why. Once the chart points at a step, switch to session replay, heatmaps, form analytics or call notes to find the cause. Teams that stop at the drop-off percentage and skip the why never actually fix anything.

A worked example: reading the sample funnel above

Run the two rules on the sample table. By percentage, the worst step is visitors to leads at 12%, an 88% drop. By absolute count, that same step also sheds the most people, 8,800, so here percentage and volume agree and the top of funnel is the priority. That is the first place to point diagnosis.

Now check the design floor. An 88% visitor-to-lead drop is normal for broad traffic, so the raw percentage alone does not prove a problem; you compare it to the 96 to 98% drop typical for cold visitors and find 88% is actually above average. The real underperformer is the closed-deal step at 35%, below a healthy 40 to 50% for warm proposals. Small band, big money. The chart made the top look scary and hid the step that a benchmark reveals. That gap between what the shape shows and what benchmarks say is the point of reading a funnel deliberately instead of eyeballing the widest gap.

This worked read is original to this guide: most funnel-chart articles stop at drawing the shape and never show you a case where the visually biggest drop is the wrong thing to fix.

Where funnel visualization fits your reporting

A funnel chart is one panel in a wider view, not the whole dashboard. Pair it with trend lines for each conversion rate over time and a source breakdown so you can see whether a leak is universal or channel-specific. See how it sits beside the other panels in our guide to the eight charts every marketing dashboard needs.

If your funnel view keeps surfacing the same leak and you want a partner to design the tracking and fix the drop, book a consultation and we will build the funnel reporting and the repair plan together.

Frequently asked questions

What is funnel visualization?

Funnel visualization is turning your stage-by-stage counts into a chart shaped like a funnel, wide at the top and narrow at the bottom, so the number of people surviving each step is shown by band width. Its job is to make the biggest drop-off between two stages visible at a glance, without you having to compare raw numbers by hand.

How do I make a funnel chart in Excel?

Put your stage names in one column and their counts in the next, in top-of-funnel order. Select both columns, click Insert, then the Waterfall, Funnel, Stock, Surface or Radar button, then choose Funnel. Excel draws a centered horizontal funnel automatically with no formulas. Add a percentage column beforehand if you want proportion labels alongside the raw counts.

Can I build a funnel chart in GA4?

Yes. Open Explore in GA4, pick the Funnel exploration template, and define each step as an event or condition in sequence. GA4 returns a stepped-bar funnel plus a table showing user counts and drop-off rate at every step. Use it when your stages are on-site events like sign-ups or purchases; use Looker Studio instead when the funnel spans ads, CRM and analytics together.

How do I read drop-off in a funnel chart?

Read both the percentage and the absolute count. The step-to-step percentage flags the leakiest step, but the raw number of people lost tells you which leak is worth fixing first, because a smaller percentage on a huge volume can outweigh a big percentage on a tiny one. Then compare each step to its own benchmark, since some steps are designed to shed most people.

How many stages should a funnel chart have?

Keep it to four to seven stages. Fewer than four and the chart tells you nothing a single sentence could not. More than seven and the drop between adjacent bands gets too thin to act on, so real leaks hide inside the noise. If your true funnel has more steps, collapse the minor ones and chart only the stages where a decision actually changes.

What is the difference between a funnel chart and a bar chart?

A bar chart compares independent categories that have no required order. A funnel chart shows a dependent sequence where each stage can only draw from the stage above it, so counts can only stay level or fall down the funnel. If a lower band ever renders wider than the one above it, either your steps are out of order or the data is wrong.