Funnel Chart in Excel — Free Template + Step-by-Step Guide (2026)

Every pipeline leaks — the question is where. A funnel chart shows each stage of a process as a shrinking bar, making the biggest drop-off impossible to miss. In 35 minutes you’ll build one in Excel two ways: the one-click built-in funnel chart (Excel 2019+/365) and the classic centred stacked-bar funnel that works in any version — plus conversion-rate columns, proper labels, drop-off analysis, and an in-cell data-bar alternative for dashboards. Works in Excel 2010 through Microsoft 365.

Dr Waqas Rafique · Founder & CTO · PhD, Statistical Machine Learning
LinkedIn · About

What is a Funnel Chart?

A funnel chart visualises a multi-stage process where quantity shrinks at every step. Each stage is a horizontal bar whose width is proportional to its count, the bars are centred, and they stack top-to-bottom from widest to narrowest — producing the shape of, well, a funnel. The classic example is a sales pipeline: 10,000 website visitors become 2,400 leads, 960 marketing-qualified leads, 380 sales-qualified leads, and finally 95 customers. One glance at the funnel and you see both the overall narrowing and exactly which transition bleeds the most.

That last part is the chart’s real job. The absolute numbers matter less than the ratios between adjacent stages — the conversion rates. A funnel whose third bar is dramatically narrower than its second is shouting that the MQL→SQL handoff is the leak. A table of five numbers contains the same information, but the funnel shape makes the comparison pre-attentive: your eye finds the pinch point before your brain reads a single digit. That is why funnels dominate marketing dashboards, sales reviews, recruitment pipelines, and product-onboarding analyses.

Excel’s support comes in two tiers. Excel 2019, 2021, and Microsoft 365 include a native Funnel chart type — select your data, click, done. Excel 2016 perpetual and earlier do not, but the workaround — a stacked bar chart with an invisible “spacer” series that pushes each bar into the centre — is a ten-minute build that has served analysts for two decades and is arguably more customisable than the native chart. This guide covers both, because the version question is the first thing every funnel-chart search runs into.

One caveat worth stating early: a funnel chart describes counts at stages, not time or causality. It will not tell you how long prospects sit in each stage or why they leave — for that you pair it with the drop-off table and pipeline-velocity metrics covered in Steps 6 and 8.

Before You Start

Funnel data looks trivial — five labels and five numbers — but two definitional choices decide whether the chart tells the truth.

Decide: cohort funnel or pipeline snapshot

A cohort funnel follows one group through every stage: of January’s 10,000 visitors, how many eventually became leads, MQLs, and customers? Counts must shrink monotonically and conversion rates are true probabilities. A pipeline snapshot counts whoever is sitting in each stage right now, regardless of when they entered — useful for capacity planning, but stage counts can legitimately exceed the one above, and “conversion rates” computed from a snapshot mix cohorts and mislead. Both are valid charts; label which one you are building.

Use consistent entities

Every stage must count the same kind of thing. Mixing “sessions” at the top with “people” in the middle and “deals” at the bottom quietly breaks the ratios. Pick one entity (people or accounts, usually) and define each stage as a filter on it.

Five to seven stages, maximum

Funnels with ten stages turn into stripes nobody reads. Collapse minor sub-stages; the chart should show the transitions you can act on, and the detail can live in the source table.

1.Step 1 — Lay Out the Funnel Data

Two columns, widest stage first. Use this layout (it is also the worked-example data for the rest of the guide):

Order matters: Excel draws the funnel in row order, top row at the top. If your source data arrives in another order, sort it before charting — or, in a cohort funnel where counts must shrink, sorting by count descending produces the same result.

If you are pulling from raw records, build each count with a formula so the funnel stays live:

Here column D on the CRM sheet holds the furthest-stage-reached for each contact, column B the cohort date, and E1 the period you are analysing. Counting “furthest stage reached” (rather than “current stage”) is what makes it a true cohort funnel.

2.Step 2 — Insert the Built-in Funnel Chart (Excel 2019+/365)

If you are on Excel 2019, 2021, or Microsoft 365 (including Excel for the web), the funnel is a native chart type:

1. Select A1:B6 — both columns, headers included.

2. Go to Insert → Recommended Charts → All Charts → Funnel. (Shortcut: the small Waterfall/Funnel icon in the Charts group on the Insert ribbon also lists Funnel directly.)

3. Click OK. Excel draws five centred horizontal bars, widest at the top, each labelled with its value.

That is genuinely all. The native chart handles the centring, ordering, and labelling that older versions needed a workaround for. A few worthwhile tweaks straight away: delete the chart title placeholder or type a real one (“Q2 Acquisition Funnel”); click any bar and use Format Data Series to set the gap width around 10–15% so the bars read as a connected funnel rather than floating stripes; and resize the chart taller than wide — funnels read vertically.

The honest limitations of the native chart: it labels raw values only (percentages need the helper-label trick in Step 5), every bar must be one series (no per-stage colour control beyond manual point formatting), and it does not exist in Excel 2016 perpetual or earlier — which is why Step 3 exists.

3.Step 3 — The Stacked-Bar Funnel for Older Excel

The classic workaround builds the funnel from a horizontal stacked bar chart plus an invisible “spacer” series that pushes each visible bar into the centre. Insert a helper column between stages and counts (so counts now sit in column C), and in B2 enter:

Fill down. Each spacer is half the difference between that stage and the widest stage — exactly the left-margin needed to centre the bar. For Visitors (the max) the spacer is 0; for Customers it is (10,000−95)/2 = 4,952.5.

Now build the chart:

1. Select A1:C6 and insert a Stacked Bar chart (the horizontal one, second row of the Bar gallery).

2. Click the spacer series (the left segments) → Format Data Series → Fill: No fill, Border: No line. The spacers vanish, and each count bar now floats centred.

3. Excel plots the first row at the bottom of a bar chart, so the funnel is upside-down. Double-click the vertical axis → Axis Options → tick Categories in reverse order. Widest stage jumps to the top.

4. Tidy: delete the legend (the spacer entry is meaningless), delete the horizontal value axis, and set the series Gap Width to ~10% so the bars nearly touch.

The result is visually indistinguishable from the native funnel — and because each half of the build is an ordinary chart element, you can format it more freely: per-bar colours, gradient fills, even a second mirrored spacer for asymmetric designs.

4.Step 4 — Add a Conversion-Rate Column

The funnel shape shows that the pipeline narrows; the conversion columns say by how much. Add two columns beside the counts. First, stage-to-stage conversion — each stage divided by the one directly above. In D3 (beside Leads):

Fill down to D6 and format as a percentage. With the example data: Leads 24.0%, MQL 40.0%, SQL 39.6%, Customers 25.0%. Leave D2 blank — the top stage converts from nothing.

Second, overall conversion — each stage divided by the very top, with the denominator locked. In E3:

Fill down: 24.0%, 9.6%, 3.8%, and finally 0.95% of visitors become customers. The two columns answer different questions. Stage-to-stage finds the leak (which transition underperforms); overall states the end-to-end efficiency that finance and growth models need — nearly one customer per hundred visitors, here.

These percentages are also where benchmarking lives. Comparing your 24% visitor→lead rate to last quarter’s 31% is actionable; comparing your funnel’s shape to another company’s is mostly noise, because stage definitions differ everywhere.

5.Step 5 — Labels & Formatting

A funnel chart presents in meetings, so the formatting pass matters more than usual.

Show values and percentages together

The native funnel labels values only. To get “Leads — 2,400 (24%)” style labels, build a helper label column and chart it as the category instead of the plain stage names:

Because the text is built by formula, the labels update with the data — no manual re-typing when March numbers replace February’s.

Colour with intent

Use a single hue stepping from dark at the top to light at the bottom (or your brand gradient). Avoid the default rainbow: five unrelated colours imply five unrelated categories, but funnel stages are one population progressively filtered. Reserve a contrasting colour for one thing only — highlighting the problem stage you want the room to discuss.

Strip the clutter

Delete gridlines, the value axis (the labels carry the numbers), and the legend. Set the chart background to none if it sits on a coloured dashboard. A funnel needs exactly three visual elements: bars, labels, title.

6.Step 6 — Stage-to-Stage Drop-Off Analysis

The funnel shows the survivors; drop-off analysis counts the casualties. Add two more columns. The drop count — how many were lost at each transition — in F3:

And the drop rate — the share of the incoming stage that was lost — in G3:

With the example data: Visitors→Leads loses 7,600 (76%), Leads→MQL loses 1,440 (60%), MQL→SQL loses 580 (60%), SQL→Customers loses 285 (75%). Two different rankings emerge — and this is the analytical heart of the funnel. By count, the top of the funnel dominates (7,600 lost visitors). By rate, the bottom transition (75% of qualified opportunities failing to close) may be the more fixable scandal, because those are prospects who already cleared three filters.

Prioritise with leverage in mind: improvements multiply through the funnel. Lifting SQL→Customer from 25% to 30% raises customers from 95 to 114 — a 20% revenue lift from one stage — while the same five points added to visitor→lead conversion mostly adds low-intent leads that die downstream anyway. The drop-off table, not the funnel picture, is what makes that argument with numbers.

7.Step 7 — The Conditional-Formatting Funnel

Sometimes a full chart is overkill — a dashboard cell, a weekly email table, a sidebar widget. Excel’s Data Bars turn the count column itself into a funnel:

1. Select the counts C2:C6.

2. Home → Conditional Formatting → Data Bars → pick a gradient or solid fill.

3. Each cell now contains a bar proportional to its value — an instant left-aligned funnel living next to the numbers and the conversion columns.

Two refinements lift it from default to deliberate. First, open Conditional Formatting → Manage Rules → Edit Rule and set Minimum to Number: 0 and Maximum to a fixed number a little above your top stage — otherwise Excel rescales the bars to the current min/max and this month’s funnel is not visually comparable to last month’s. Second, tick Show Bar Only in a duplicated column if you want a clean bar lane beside an untouched number lane.

The data-bar funnel is left-aligned rather than centred — technically a “decay bar chart” — but for analysis that is arguably better: aligned left edges make length comparisons more accurate than centred symmetry, at the cost of looking less like the marketing-deck funnel icon. Use the centred chart to present, the data bars to monitor.

8.Step 8 — Common Variants

Period-over-period comparison funnels

Place this month’s and last month’s funnels side by side on the same value scale (fix both charts’ max via the spacer reference or axis settings). Better still for analysis: chart the conversion rates for both periods as a clustered bar — shape comparisons hide rate changes, rate charts expose them.

Pipeline-velocity table

A funnel counts; it does not time. Pair it with average days-in-stage per transition (e.g. =AVERAGEIFS(DaysInStage,Stage,"SQL")). A stage can convert well but slowly — a different disease with a different cure, invisible in the funnel itself.

Weighted (revenue) funnels

Replace counts with the sum of deal values per stage (=SUMIFS(Value,Stage,A2)). The shape often changes dramatically — a few large deals can make a numerically thin stage financially fat — and the revenue funnel is usually what the board actually wants.

Inverted funnels

Growth processes that widen — one piece of content → impressions → shares → sign-ups from shares — can be drawn as an upside-down funnel (smallest first). The native chart will happily render increasing values; just order the rows smallest-first and the “funnel” becomes a megaphone. Use sparingly: audiences expect funnels to narrow, so label the inversion loudly.

Try It: Live Funnel Builder

Type your own stage counts below — visitors through customers — and watch the centred funnel redraw live, with stage-to-stage conversion percentages on every transition and the leak (worst conversion) flagged automatically. This is exactly what your Excel chart from Step 2 or Step 3 produces, plus the Step 4 percentages baked in.

Advanced Funnel Tips

Segmented funnels with COUNTIFS

One funnel hides segment differences. Build the count formulas with an extra criterion — =COUNTIFS(Stage,A2,Channel,$H$1) — and drive H1 from a Data Validation dropdown of channels. Flipping the dropdown redraws the funnel per channel in place: paid search may convert visitors to leads at triple the rate of social, a fact the blended funnel averages away.

Funnel + cohort: the retention bridge

The funnel ends at “customer,” but the business does not. Feed your customer counts into a cohort retention matrix to see how each month’s funnel output survives over time — acquisition quality and retention quality are different axes, and great funnels sometimes produce churn-prone customers.

Confidence on small numbers

With 95 customers, a swing of ±10 is sampling noise, not a trend. Rough rule: the month-to-month noise on a conversion rate is about ±1/√n of the incoming count — on 380 SQLs that is ±5 points. Resist re-engineering the sales process over rate changes inside the noise band; track a 3-month moving average of each rate instead.

From monthly screenshot to live funnel

A funnel rebuilt by hand every month is a chore that dies the first busy week. Upload your pipeline export to DataHub Pro and it maintains the funnel, the conversion rates, and the period comparisons automatically — one shareable link that is always current, instead of a deck of stale screenshots.

Worked Example: Finding a SaaS Funnel’s Leak

A B2B SaaS startup runs the funnel from this guide: 10,000 monthly visitors → 2,400 leads → 960 MQLs → 380 SQLs → 95 customers. The marketing lead builds the table in Step 1, the native funnel chart in Step 2, and the conversion columns in Step 4: 24% → 40% → 39.6% → 25%, with overall visitor→customer conversion at 0.95%.

The funnel picture draws every eye to the top — the dramatic 76% narrowing from visitors to leads. The CEO’s instinct is “we need more traffic and better landing pages.” But the Step 6 drop-off table reframes it: a 24% visitor→lead rate is healthy for the sector, while the 25% SQL→customer close rate sits well under the 30–35% the sales team hit last year. 285 sales-qualified opportunities — prospects who survived three filters and took a demo — are dying at the final step each month.

The leverage maths from Step 6 settles the argument. Restoring close rate to 32% lifts customers from 95 to 122 — +28% revenue — with zero extra marketing spend. Achieving the same gain at the top would require ~2,800 additional monthly visitors of equal quality. The team A/B tests the post-demo follow-up sequence and adds a pricing-objection playbook; two months later the funnel redraws with a visibly fatter bottom bar, and the segmented version (Advanced tips) shows the gain concentrated in the mid-market channel.

The lesson generalises: the funnel chart locates the conversation, the conversion and drop-off columns decide it. Build all three layers every time — the chart alone optimises for whichever bar shrinks most photogenically, which is rarely the most profitable fix.

Common Errors & Fixes

“Funnel” is missing from the chart gallery

You are on Excel 2016 perpetual or older — the native funnel ships with 2019+/365 only. Build the stacked-bar funnel from Step 3; the output is equivalent.

The stacked-bar funnel is upside-down

Bar charts plot the first row at the bottom. Double-click the category axis → Axis Options → tick Categories in reverse order. If the horizontal axis then jumps to the top, that is normal — delete it anyway per Step 3.

The bars are not centred

The spacer series is missing, in the wrong column order (spacer must be plotted before the count series), or still visible. Check the formula is =(MAX($C$2:$C$6)-C2)/2 with the count range absolute, and the series order in Select Data shows spacer first.

A conversion rate is over 100%

A stage is larger than the one above — you are charting a pipeline snapshot, mixing cohorts, or counting different entities per stage. Re-define every stage as “members of cohort X who reached at least this stage.”

#DIV/0! in the conversion column

An upstream stage is zero or blank. Guard the formula: =IF(C2=0,"",C3/C2) — common when a new period’s data has not loaded yet.

The funnel looks different every month even with similar rates

Conditional-formatting data bars are auto-rescaling, or the chart axis max is floating. Fix the data-bar maximum to a constant in Manage Rules, and fix comparison charts to a shared scale.