PlotSet LogoPlotSet
All chart types

Scatter Plot Maker

Plot two variables against each other and see whether they actually move together. Upload a spreadsheet, pick your X and Y columns, and publish an interactive scatter plot you can embed anywhere — free.

Create a Scatter Plot — freeSee the gallery

What is a Scatter Plot?

A scatter plot places one dot per record on two numeric axes — one variable on X, one on Y. Nothing is aggregated and nothing is summed: every row in your data becomes a mark you can see. That makes it the fastest way to answer "do these two things move together?", and the only common chart that shows you the shape of that relationship rather than a single summary number. Correlation coefficients hide clusters, curves and outliers; a scatter plot shows all three at a glance.

When to use a Scatter Plot

  • You are testing a relationship between two numbers

    Ad spend against revenue, price against rating, house size against sale price. If both variables are numeric and you want to know how one tracks the other, this is the default chart.

  • You need to find outliers

    Points far from the crowd are obvious on a scatter plot and invisible in a table or a bar chart. Analysts use them for exactly this: spotting the record that breaks the pattern.

  • You want to compare groups on the same two axes

    Colour the dots by category — region, segment, asset class — and you can see whether the groups occupy different parts of the space or overlap completely.

  • Your data has too many rows to plot as bars

    A bar chart with 400 categories is unreadable. A scatter plot with 400 points reads fine, because the eye reads the cloud rather than each mark.

When not to use a Scatter Plot

  • One of your variables is a category

    Countries, products and departments have no numeric position on an axis. Use a bar chart for category-against-value comparisons, or a box plot if you want the distribution within each category.

  • You are showing change over time

    Time on X with values on Y is a line chart. Points alone drop the connection between consecutive periods, which is the thing you are trying to show.

  • You have fewer than about ten records

    With a handful of points there is no cloud to read and no pattern to trust. Show the numbers directly in a table or a labelled bar chart.

  • You want to prove causation

    A scatter plot can show that two variables move together. It can never show that one causes the other, and presenting it as if it does is the single most common misuse of the chart.

How to make a Scatter Plot

From a spreadsheet to a published chart, step by step.

01

Start from the scatter plot template

Open the PlotSet scatter plot template. It ships with sample data already bound, so you can see the shape of the spreadsheet you need before you upload your own.

02

Bring in your data

Upload a CSV or Excel file, paste from a spreadsheet, or connect a Google Sheet. You need at least two numeric columns — one per axis — plus an optional label column for tooltips and an optional category column for colour.

03

Bind your columns to the axes

Assign one numeric column to X and one to Y. Add a category column to colour and group the points, and a size column if you want the marks scaled by a third variable.

04

Label the axes and the outliers

Give both axes real names with units. Turn on point labels for the records that matter — the outliers usually carry the story — and leave the rest to the tooltip so the chart stays readable.

05

Publish, embed or export

Publish to get a live interactive embed for your site or blog, or export a PNG, SVG or PDF for a slide deck or report.

Start from the Scatter Plot template with sample data already loaded, then swap in your own.

Open the template

The parts of a Scatter Plot

X axis
The horizontal scale, carrying the variable you treat as the input or the explanation — spend, size, time invested.
Y axis
The vertical scale, carrying the variable you are trying to explain — revenue, price, score.
Data point (mark)
One dot per record. Its position encodes two values at once, which is why a scatter plot holds more information per pixel than almost any other chart.
Colour / grouping
An optional category encoded as colour, letting you compare several groups on one pair of axes without splitting into separate charts.
Trend line
An optional fitted line summarising the direction and strength of the relationship. Useful as a guide, dangerous when the underlying cloud is curved or clustered.
Outlier
A point far from the main cloud. Worth labelling directly — it is usually the most interesting record on the chart.

Common Scatter Plot mistakes

Truncating the axes to exaggerate a pattern

Cutting the axis range zooms into noise and makes a weak relationship look tight. Start from a range that reflects the real spread of the data, and say so if you deliberately zoom.

Overplotting

With thousands of points the cloud turns into a solid blob and the density information is lost. Reduce the mark size, add transparency, or aggregate before plotting.

Adding a trend line to data that is not linear

A straight line through a curved or clustered cloud reports a relationship that is not there. Look at the shape first, then decide whether a summary line is honest.

Leaving the axes unlabelled

A scatter plot is meaningless without units. Two unlabelled numeric axes are the fastest way to make a correct chart unreadable.

Scatter Plot: common questions

Showing the relationship between two numeric variables. Each record becomes one point positioned by both values, so you can see whether the two move together, how tightly, and which records break the pattern.

Make your Scatter Plot now

Free to start with 300+ templates. No credit card required.

Create a Scatter Plot — freeAll chart types