| Stage | Count |
|---|---|
| Visited Pricing Page | 18400 |
| Started Trial | 4120 |
| Activated Account | 2380 |
| Added Payment Method | 910 |
| Became Paying Customer | 640 |
How to Create a Funnel Chart
A step by step walkthrough for turning a stage-by-stage spreadsheet into a published funnel chart in PlotSet, from ordering the rows to exporting or embedding the result.
Etwa 7 MinutenApplicants to hires, visitors to purchasers, leads to closed deals: how to create a funnel chart in PlotSet works the same way regardless of which stage counts are narrowing. You will upload the sheet, get the stage order right, label the drop-off between bands, and either publish a live embed or export a static image for a deck.
If your data is already two columns, stage and count, in the right order, this takes about five minutes. Add a couple more if you need to reconcile counts that came from different reports, since a funnel only works when each stage is genuinely a subset of the one before it. Budget the seven minutes below and treat the rest as buffer for that cleanup.
Die Daten, die Sie brauchen
Two columns, one row per stage, already sorted in process order: the stage name first, the count still remaining second. No totals row, one header row.
Row order is chart order. PlotSet draws the bands top to bottom in the order they appear in the file, so sort your rows by process stage before uploading, not by count.
Every stage must be a subset of the one above it. If some paying customers never technically "started a trial" in your tracking, the counts will not nest cleanly and the chart will misrepresent the process. Fix the definitions in the source data first.
Keep it to three to six stages. Below three you just have a conversion rate, and past six the bands get too thin to label.
The count column must decrease (or stay flat) from row to row. A stage that goes back up usually means two different metrics got mixed into one column, for example a raw pageview count sitting above a deduplicated visitor count.
Die Schritte
Der genaue Klickpfad in PlotSet, von einer leeren Vorlage bis zum fertigen Diagramm.
- 01
Open the funnel chart template
From the PlotSet dashboard, start a new project and choose the Funnel Chart template. It loads with sample stages already bound to the columns, so you can see the two-column shape before touching your own file.

Jede Anleitung beginnt hier. Wählen Sie erst die Kategorie, dann die Vorlage. Das Raster ist für jeden Diagrammtyp dasselbe. - 02
Order your rows by stage, not by count
In your spreadsheet, put the first stage of the process in row one and work down to the last. This is the step people skip, and it is the one PlotSet has no way to guess for you: nothing about the numbers themselves tells the chart which stage came first.
- 03
Bring in the sheet or connect it live
Import a CSV or .xlsx file from the data panel, paste rows directly from a spreadsheet, or connect a Google Sheet if the counts get refreshed on a schedule and you want the chart to update with them.

Die Datenansicht, hier in einem Balkendiagramm-Projekt. Laden Sie über "Upload CSV File" eine CSV-Datei hoch, holen Sie die Zahlen mit "Magic Import" aus einem Bild oder verbinden Sie ein Google Sheet. Der Bereich sieht bei jedem Diagrammtyp gleich aus. - 04
Map the stage and value columns
Confirm which column feeds the stage labels and which feeds the band width. PlotSet usually gets this right from clear headers like "Stage" and "Count", but check it if your sheet carries extra columns it might have guessed wrong on.

Die Zuordnungen legen fest, welche Spalte zur Beschriftung wird, welche die Werte liefert und welche die Farbe steuert. "Auto Binding" errät sie anhand Ihrer Spaltenüberschriften. Prüfen Sie das Ergebnis, statt sich darauf zu verlassen. - 05
Turn on stage-to-stage conversion labels
Switch on the label option that shows the percentage carried over from the previous stage, not just the raw count at each one. The count alone tells you the size of each stage; the conversion percentage tells you where the process is actually losing people.
- 06
Style the bands
A single colour, darkening slightly from top to bottom, reads as one continuous process rather than a set of unrelated categories. Reserve a second, contrasting colour for the one stage you want the reader to notice, usually the biggest drop.

Der Tab "Preview" für ein Diagramm vom Typ "Funnel Chart". Alle Gestaltungsoptionen stecken im umrandeten Bereich: Aussehen, Kopfzeile, Fußzeile und die Optionen, die es speziell für diesen Diagrammtyp gibt, jeweils in einem eigenen Tab. - 07
Title the chart around the finding
"Trial signup funnel" describes the data; "Most trials drop off before adding a payment method" states the finding. Put the second kind of title on the chart and let the bands back it up.
- 08
Publish, embed or export the funnel
Publish for a live interactive chart where hovering a band shows its exact count and conversion rate, grab the embed code for a page, or export a PNG, SVG or PDF for a slide or report.
Starten Sie mit der Funnel Chart-Vorlage samt bereits geladenen Beispieldaten und ersetzen Sie diese durch Ihre eigenen.
Vorlage öffnenStyling-Tipps für Funnel Chart
Label the drop-off between bands, not just each band
A funnel chart with only the count at each stage asks the reader to do subtraction in their head. Show the amount or percentage lost between consecutive stages directly on the taper, so the worst step is visible without any arithmetic.
Keep the taper, skip the 3D effect
Some tools offer a beveled or 3D-shaded funnel. It looks like a decoration from a slide template and makes the band widths harder to compare accurately. A flat funnel with straight taper edges reads faster and measures better.
Order stages by process, never by size
It is tempting to sort by count when a report happens to come out that way, but a funnel's row order is read as "what happens first," not "what is largest." If two stages could ever swap order and still make sense, they do not belong in the same funnel.
Match colour intensity to stage, not to value
A single hue that gets slightly darker toward the bottom mirrors the shrinking shape and needs no legend. Colouring bands by their count instead just repeats information the width already shows.
Häufige Fehler bei Funnel Chart
The bands are not actually narrowing in order
This almost always means the rows were sorted by count, alphabetically, or by whatever order they came out of a report, not by the order the stages happen in. Re-sort the source rows by process stage in the spreadsheet and re-upload; PlotSet renders top to bottom in row order and cannot infer the process for you.
One band is wider than the stage above it
A later stage cannot legitimately have more people than an earlier one in a true funnel, so this points to two different counting methods getting mixed together, for example a session count above a unique-user count. Check how each number was produced and make sure every stage counts the same population, just filtered further.
The chart has too many thin, unreadable bands
Past about six stages, minor sub-steps crowd out the ones that matter and every band gets too thin to label. Group the intermediate steps into the handful of stages someone could realistically act on, and move the finer breakdown into a table alongside the chart.
Mehr über Funnel Chart erfahren
Nicht sicher, ob Funnel Chart die richtige Wahl ist? Erfahren Sie, was es ist, wann Sie es verwenden sollten und wann nicht.
Funnel Chart: häufige Fragen
Two columns and one header row, with the stage name in the first column and the count remaining at that stage in the second. Rows must already be sorted in process order, since PlotSet draws bands top to bottom in the order they appear in the file.
Jetzt Funnel Chart erstellen
Kostenlos starten mit über 300 Vorlagen. Keine Kreditkarte erforderlich.



