January 5th, 2026
When To Use A Scatter Plot: Examples & Limitations in 2026
By Zach Perkel · 26 min read
After creating hundreds of data visualizations, I know that choosing when to use a scatter plot can make or break your analysis. Here's how to decide if a scatter plot is the right fit for your data in 2026.
What is a scatter plot?
A scatter plot is a chart that displays two numerical variables as dots on an X and Y axis, where each dot represents one data point. The position of each dot shows how the two variables relate to each other, making it easy to spot patterns like positive correlations (both variables increase together), negative correlations (one rises as the other falls), or no relationship at all.
I use scatter plots when I need to answer questions like "Does spending more on ads lead to more sales?" or "Is there a connection between employee tenure and performance ratings?" They're also invaluable for catching outliers that would be hard to see in a table of numbers.
When to use a scatter plot: Uses and applications
Scatter plots are the right choice when you need to understand how two variables relate to each other. They answer questions like "Does X affect Y?" and help you spot patterns that tables and bar charts would hide. Here are some common situations where a scatter plot is the best tool for the job:
Finding correlations: When you want to know if two metrics move together, a scatter plot shows the relationship visually. Plotting customer support response time against satisfaction scores reveals whether faster replies lead to happier customers.
Identifying outliers: Data points that sit far from the rest of the cluster stand out immediately on a scatter plot. These unusual values could signal data entry errors, fraud, or high-performing accounts worth investigating further. I've caught billing mistakes this way that would have taken hours to find in a spreadsheet.
Testing assumptions: Before investing in a strategy, you can use a scatter plot to check if your assumptions hold. If your team believes that longer sales calls close more deals, plotting call duration against close rate shows whether the data supports that belief. I run this kind of check before recommending any major budget changes.
Comparing performance across segments: Scatter plots help you see how different groups stack up. Plotting revenue against customer acquisition cost for each marketing channel shows which channels deliver the best return.
Spotting trends over two dimensions: When you need to see how two metrics interact rather than how one changes over time, a scatter plot captures that relationship. Plotting employee training hours against productivity scores shows whether more training correlates with better output.
Exploring data before deeper analysis: Scatter plots are useful for early-stage exploration when you're not yet sure what story the data tells. A quick plot can reveal patterns worth investigating or confirm that two variables have no meaningful connection. I typically start with a simple scatter plot before moving on to more complex analysis.
Running regression analysis: Regression analysis measures how strongly two variables are connected and helps predict one based on the other. A scatter plot is the starting point because it shows whether a linear relationship exists. If you're trying to forecast revenue based on lead volume, plotting both variables first tells you whether a prediction model makes sense.
8 Types of scatter plots
Scatter plots come in several formats, and each one answers a slightly different question. A simple scatter plot tells you whether two variables are related, while a bubble chart can show three variables at once. Here are the main types you'll encounter:
1. Simple scatter plot
A simple scatter plot shows the relationship between two variables with plain dots on a grid. This is the most common type and works well when you want a quick look at whether two things are connected.
For example, plotting customer satisfaction scores against repeat purchase rates shows whether happier customers buy more often. I reach for this version first when I'm exploring data and don't yet know what I'm looking for.
2. Scatter plot with trend line
Adding a trend line (also called a line of best fit) helps you see the overall direction of your data. The line cuts through the middle of your dots and shows whether the relationship trends upward, downward, or stays flat. If you're plotting advertising spend against website traffic, a trend line instantly shows whether more spending generally leads to more visitors.
I find this version helpful when presenting to stakeholders who want a clear takeaway from the data without studying individual dots.
3. Bubble chart
A bubble chart is a scatter plot where the size of each dot represents a third variable. Instead of uniform dots, you get circles of different sizes that add another layer of information. For instance, you could plot marketing spend versus revenue, with bubble size showing the number of customers in each region.
I recommend this type when you need to compare three dimensions at once, but don't want to overwhelm your audience with multiple charts.
4. Scatter plot with categories
This type uses different colors or shapes to separate data points into groups. If you're comparing sales performance across multiple product lines, each line gets its own color. This makes it easy to spot whether patterns differ between categories or if one group behaves differently from the rest.
In my experience, color-coded scatter plots are the fastest way to answer questions like "Is this trend consistent across all our segments?"
5. 3D scatter plot
A 3D scatter plot adds a third axis to display three variables in three-dimensional space. For example, you could plot sales revenue on one axis, marketing spend on another, and customer acquisition cost on the third to see how all three relate.
While this sounds useful, I rarely recommend it for business presentations. The added dimension can make patterns harder to read on a flat screen, and rotating the chart to find the right angle gets frustrating. In most cases, a bubble chart communicates the same information more clearly.
6. Scatter plot matrix
A scatter plot matrix (sometimes called a SPLOM) arranges multiple scatter plots in a grid so you can compare several variables at once. Each cell in the grid shows a different variable pairing.
Say you have revenue, customer count, average order value, and return rate in one dataset. A matrix displays how each variable relates to the others without building separate charts. This type is useful during early data exploration when you're not sure which relationships matter yet. I use it to scan for patterns quickly before building a more focused chart.
7. Connected scatter plot
A connected scatter plot draws lines between dots in sequence, usually to show change over time. Unlike a standard line chart, it keeps both variables visible rather than assuming time is one of them. This works well for tracking how two metrics evolve together, like plotting monthly ad spend against monthly revenue over a year.
8. Density scatter plot
When you have thousands of data points, a standard scatter plot turns into a blob of overlapping dots. A density scatter plot (or hexbin plot) solves this by using color shading or hexagonal bins to show where points cluster. Darker areas mean more data points.
If you're analyzing transaction data from 50,000 customers, this type shows you where most purchases fall in terms of price and quantity without losing the pattern in visual clutter. I turn to this type when working with large datasets where individual dots would be impossible to read.
How to read scatter plots
Reading a scatter plot takes just a handful of steps once you know what to look for. Here's how to pull insights from any scatter plot you encounter:
Start with the axes: Before looking at the dots, check what each axis represents. The horizontal axis (X) shows one variable, and the vertical axis (Y) shows the other. Make sure you understand the units too. If the X axis shows "monthly ad spend" and the Y axis shows "new customers," you're looking at how spending relates to customer acquisition.
Look at the overall pattern: Step back and observe the general shape the dots form. If they slope upward from left to right, there's a positive correlation, meaning as one variable increases, so does the other. If they slope downward, there's a negative correlation. If the dots scatter randomly with no clear direction, the two variables probably aren't connected.
Check how tightly the dots cluster: A tight cluster along a line means a strong relationship. A loose, spread-out pattern means the relationship is weak, even if a general trend exists. I pay close attention to this because a weak correlation can lead to overconfident decisions if you don't notice the spread.
Spot the outliers: Look for dots that sit far away from the main group. These are your unusual values. Ask yourself why they're different. Is it a data error? An unusual circumstance? A top performer worth learning from? I've found that outliers often contain the most valuable insights.
Read the trend line if there is one: If the scatter plot includes a trend line, use it to see the average direction of the relationship. The steeper the line, the stronger the effect of one variable on the other. A flat line means changes in X don't really affect Y.
How to create a scatter plot in Julius
Creating a scatter plot in Julius doesn't require writing code or building charts manually. You type a question in plain English, and Julius builds the chart for you. Here's how to do it:
Connect your data: Link Julius to where your data lives. This could be a database like Postgres, Snowflake, or BigQuery, or a simple file like a CSV or Excel spreadsheet. Julius also connects to Google Drive, Stripe, Intercom, and Google Ads. Once connected, Julius can access your data whenever you ask a question.
Ask a question about two variables: Type a question about the two things you want to compare. Something like "show me the relationship between ad spend and conversions" or "plot customer lifetime value against purchase frequency." Julius figures out which columns to use and builds the chart automatically.
Get your scatter plot: Julius creates a scatter plot based on your question and displays it in the conversation. You can hover over any dot to see the exact values it represents.
Add details if needed: If your data has lots of columns or similar names, Julius may ask a follow-up question to make sure it's using the right information. You can also narrow your results by adding filters like "only for Q1 2024" or "exclude returns."
Save or export: Once you're happy with the chart, download it as an image, PDF, or CSV. You can also share it with your team directly in Julius, or schedule it to arrive in your inbox or Slack regularly.
Build on it in a Notebook: If you want to reuse your analysis or add more steps, move it into a Julius Notebook. Notebooks save your work so you can run the same analysis again when your data updates. I use Notebooks when I need a scatter plot refreshed weekly with the latest numbers.
Examples of scatter plots
Seeing scatter plots in action makes them easier to understand, and the right example can show you exactly when to reach for this chart type. Here are two examples where a scatter plot works better than other options:
Example 1: Marketing spend vs. revenue
This scatter plot places monthly marketing spend on the X axis and monthly revenue on the Y axis. Each dot represents one month of data.
In this example, the dots slope upward from left to right, showing a positive correlation. Months with higher marketing budgets tend to produce higher revenue.
But notice the spread. The dots don't form a tight line, which tells you marketing spend isn't the only factor driving revenue.
I use this type of chart when presenting to leadership because it answers the question "Is our marketing working?" while also showing that the relationship isn't a simple one-to-one.
Example 2: Customer satisfaction vs. churn rate
This scatter plot shows customer satisfaction scores (from survey data) on the X axis and churn rate by customer segment on the Y axis. Each dot represents a different customer group, like enterprise clients, mid-market accounts, or small businesses.
The dots slope downward from left to right, showing a negative correlation. Segments with higher satisfaction scores have lower churn rates.
This chart helped me identify which customer groups needed attention. The segments sitting in the upper-left corner (low satisfaction, high churn) became immediate priorities for our customer success team.
Benefits of scatter plots
Scatter plots are a go-to chart for business analysis because they answer a specific question other charts can't: “How do two variables relate?” Here are their main benefits:
Show relationships at a glance: A scatter plot reveals whether two variables are connected without requiring any calculations. You can spot a positive or negative correlation the moment you look at the chart.
Handle large amounts of data: Unlike bar charts or tables, scatter plots can display hundreds of data points without becoming unreadable. Each dot takes up minimal space, so you can see the full picture of your dataset.
Expose outliers immediately: Data points that sit far from the main cluster jump out visually. These unusual values often represent your biggest problems or your biggest opportunities.
Support better decision-making: When you can see the strength of a relationship (tight cluster vs. loose spread), you can make more informed choices about whether to act on it. A weak correlation looks different from a strong one, and that distinction matters.
Work across departments: Marketing, finance, sales, and operations teams all deal with questions about how two metrics relate. Scatter plots translate across business functions because the visual language is intuitive.
Serve as a starting point for deeper analysis: A scatter plot often reveals patterns worth investigating further. I regularly use them as a first step before running regression analysis or building predictive models.
Limitations of scatter plots
Scatter plots are useful, but they come with some drawbacks you should know before relying on them. Here are the main limitations:
Overplotting with large datasets: When you have thousands of data points, the dots overlap and turn into a blob. You lose the ability to see patterns or individual values. A density scatter plot or hexbin chart can help, but at that point, you're no longer working with a simple scatter plot.
Correlation doesn't mean causation: A scatter plot can show that two variables move together, but it can't tell you if one causes the other. If your chart shows that ice cream sales rise when temperatures rise, that's a correlation. It doesn't mean hot weather causes people to buy ice cream (even if that seems obvious). There could be other factors at play.
Misreading clusters: It's easy to see patterns that aren't meaningful. A tight cluster of dots can look like a strong relationship, but it could be the result of limited data or a specific subset of your dataset. I've seen teams make confident decisions based on clusters that disappeared once more data came in.
Only works with numeric data: Scatter plots require two numeric variables. If you're working with categories (like product names or regions), you'll need a different chart type or a way to convert your data into numbers first.
Hard to show more than two variables clearly: Scatter plots are built to show how two variables relate. You can add a third variable using color or bubble size, but this makes the chart harder to read. If you need to compare more than two or three dimensions, a scatter plot matrix or a different chart type works better.
Best practices when using scatter plots
Before you build or present a scatter plot, run through these quick checks to make sure your chart is accurate and useful. Here are the guardrails that will save you from common mistakes:
Label your axes clearly: Every scatter plot needs axis labels that explain what each variable represents and what units you're using. "Revenue" is vague. "Monthly revenue (USD)" is specific.
Use enough data points: A scatter plot with five dots won't reveal a reliable pattern. You need enough observations for the relationship to be meaningful. For most business questions, aim for at least 30 data points.
Check for outliers before concluding: One or two extreme values can skew the entire visual. Decide whether those outliers are errors to remove or legitimate data worth investigating separately.
Don't force a trend line onto scattered data: If your dots are spread randomly with no visible pattern, adding a trend line creates a false sense of direction. Only include a trend line when the data actually supports it.
Question the relationship before acting on it: Ask yourself if there's a logical reason these two variables would be connected. If you can't explain why they'd relate, the pattern you're seeing may be a coincidence.
Match the chart type to your data size: For datasets under a few hundred points, a simple scatter plot works well. For thousands of rows, switch to a density plot or filter your data into meaningful segments.
Keep your audience in mind: If you're presenting to people unfamiliar with scatter plots, take 30 seconds to explain how to read them. A chart only works if your audience understands what they're looking at.
Need to create scatter plots fast? Try Julius
Once you know when to use a scatter plot, the next step is building one without the manual work. With Julius, you can create scatter plots and other visualizations by asking questions in plain language.
Julius is an AI-powered data analysis tool that connects directly to your data and shares insights, charts, and reports quickly.
Here’s how Julius helps:
Quick single-metric checks: Ask for an average, spread, or distribution, and Julius shows you the numbers with an easy-to-read chart.
Built-in visualization: Get histograms, box plots, and bar charts on the spot instead of jumping into another tool to build them.
Catch outliers early: Julius highlights suspicious values and metrics that throw off your results, so you can make confident business decisions based on clean and trustworthy data.
Recurring summaries: Schedule analyses like weekly revenue or delivery time at the 95th percentile and receive them automatically by email or Slack.
Smarter over time: With each query, Julius gets better at understanding how your connected data is organized. It learns where to find the right tables and relationships, so it can return answers more quickly and with better accuracy.
One-click sharing: Turn a thread of analysis into a PDF report you can pass along without extra formatting.
Direct connections: Link your databases and files so results come from live data, not stale spreadsheets.
Ready to see how Julius can help your team make better decisions? Try Julius for free today.
Frequently asked questions
What tools can help you create scatter plots?
Excel, Google Sheets, Tableau, and Julius are the most common tools for creating scatter plots. Spreadsheet tools work well for simple charts with smaller datasets. Tableau and Power BI offer more customization for larger amounts of data. AI-powered tools like Julius let you create scatter plots by typing a question instead of manually configuring axes and data ranges.
What is the difference between a scatter plot and a line chart?
A scatter plot shows individual data points as dots to reveal relationships between two variables, while a line chart connects points in sequence to show change over time. Use a scatter plot when you want to see if two variables are correlated. Use a line chart when you want to track how one variable changes across a continuous period like days, months, or years.
Can you use a scatter plot for categorical data?
No, because scatter plots require numeric data on both axes. If you have categorical data like product names or regions, you'll need to convert those categories into numbers or choose a different chart type. Bar charts work better for comparing categories, and box plots work better for showing how a numeric variable differs across groups.
