How to Make a Scatter Plot with ChatGPT

Trying to spot a relationship between two different data points, like your Facebook Ads spend and your Shopify sales? A scatter plot is the perfect chart for the task. It instantly turns rows of spreadsheet data into a visual map showing how one variable impacts another. This guide serves as a step-by-step tutorial for creating a great-looking scatter plot with ChatGPT.

Graphed

Build AI Agents for Marketing

Build virtual employees that run your go to market. Connect your data sources, deploy autonomous agents, and grow your company.

Build Your First Agent

What Exactly Is a Scatter Plot?

A scatter plot (or scatter graph) is a type of chart that uses dots to represent the values for two different numeric variables. The position of each dot on the horizontal (x-axis) and vertical (y-axis) axis indicates values for an individual data point. The whole point is to look for a pattern and see if there's a relationship, or correlation, between the two variables.

In business, you can use a scatter plot to answer questions like:

• Does increasing our ad spend lead to more website conversions?
• Is there a relationship between a product's price and how many units we sell?
• Do sales reps who make more calls a day also close more deals?
• Is our website's page load speed related to our bounce rate?

By plotting the data, you can typically see one of three patterns emerge:

1. Positive Correlation: The dots form a pattern that goes up and to the right. This means as one variable increases, the other variable tends to increase. (e.g., More ad spend is associated with more sales).
2. Negative Correlation: The dots form a pattern that goes down and to the right. This means as one variable increases, the other variable tends to decrease. (e.g., A higher product price is associated with fewer daily sales).
3. No Correlation: The dots are scattered everywhere with no clear pattern. This suggests there is no relationship between the two variables. (e.g., The number of tweets we post has no effect on our website traffic).

Understanding these simple patterns can help you make smarter decisions instead of just guessing about what's working.

Getting Your Data Ready for ChatGPT

Before you can get ChatGPT to visualize anything, you need to give it clean and simple data to work with. The best format is a CSV (Comma Separated Values) file. It's a universal, bare-bones spreadsheet format that ChatGPT’s data analysis feature understands perfectly.

Here’s how to prep your data:

• Keep it Simple: Open your data in a spreadsheet tool like Excel or Google Sheets. Your file should contain only the two columns of data you want to compare. Any extra columns or information can confuse the analysis.
• Use Clear Headers: Make sure the very first row contains clear, simple headers for your two columns. For example, "Ad Spend" and "Website Clicks".
• One Value per cell: Each cell should have only one number. Delete anything extra in the cells, like notes, text, or multiple data types. Just clean numerical data.
• No Blank Rows: Ensure there are no empty rows in the middle of your data. ChatGPT might interpret a blank row as the end of the data set.
• Save as a CSV: In your spreadsheet program, go to "File" > "Save As" (or "Download") and choose the Comma Separated Values (.csv) format.

AI Agents for Marketing

Crash Course

Graphed

Free PDF · the crash course

AI Agents for Marketing Crash Course

Learn how to deploy AI marketing agents across your go-to-market — the best tools, prompts, and workflows to turn your data into autonomous execution without writing code.

Download the PDF

Example Data Format

Your spreadsheet should look something like this before you save it as a CSV. Nice and tidy, with clear headers and only two columns of numerical data.

Daily Ad Spend,Conversions
120,8
150,11
180,15
200,14
250,22
300,25
310,23

Creating Your Scatter Plot with ChatGPT: a Step-by-Step Guide

Once your CSV is ready, the fun part begins. To make a chart, you’ll need a ChatGPT Plus (or higher) subscription, which gives you access to the Advanced Data Analysis model. This is the version of ChatGPT that can read files, run Python code, and generate visualizations.

Step 1: Start a New Chat and Prepare Your File

Navigate to ChatGPT. Make sure you have GPT-4 selected at the top. Near the message box at the bottom, you'll see a small paperclip icon. This is how you upload your file.

Step 2: Upload Your CSV

Click the paperclip icon and select the CSV file you just prepared. Wait for it to finish uploading. Once it's attached, you can start writing your prompt.

Graphed

Build AI Agents for Marketing

Build virtual employees that run your go to market. Connect your data sources, deploy autonomous agents, and grow your company.

Build Your First Agent

Step 3: Write a Clear Prompt

Your prompt is your instruction to ChatGPT. The more specific you are, the better your first result will be. Start by telling it what kind of chart you want and which data should go on which axis.

Here are a few prompt examples you can adapt, from basic to more detailed.

Basic Prompt:

This is the simplest way to get started. Just tell it the chart type and which columns to use.

Using the attached file, create a scatter plot. Put "Daily Ad Spend" on the x-axis and "Conversions" on the y-axis.

More Detailed Prompt:

Here, we also ask for labels and a title, which makes the chart more professional right from the start.

Please generate a scatter plot from the attached CSV. The x-axis should represent 'Daily Ad Spend' and the y-axis should represent 'Conversions'. Please also add a title to the chart called "Marketing Spend vs. Daily Conversions".

Advanced Prompt (Adding a Trendline):

A trendline (or line of best fit) is a straight line that cuts through the data in your scatter plot to show the general direction or relationship. It’s incredibly useful for seeing the overall trend even if the dots are a bit scattered.

From the uploaded dataset, produce a scatter plot to visualize the relationship between 'Daily Ad Spend' and 'Conversions'. Make sure to add a trendline to the chart to help visualize the correlation.

After you submit your prompt, ChatGPT will 'think' for a moment as it writes and executes code in the background. In a few seconds, it will produce your scatter plot as an image directly in the chat.

Fine-Tuning Your Scatter Plot with Follow-Up Prompts

One of the best features of using ChatGPT is that you don't have to get it perfect on the first try. You can ask for modifications conversationally, just like you would with a human data analyst.

Let's say the first version of the chart is good, but not quite perfect. Here are some follow-up prompts you can use to tweak it.

• For Changing Colors and Styles:
• For Adjusting Labels and Titles:
• For Adding or Removing Elements:

You can go back and forth as many times as you need. This quick, iterative process is perfect for exploring your data without needing to learn any complex charting software menus or syntax.

AI Agents for Marketing

Crash Course

Graphed

Free PDF · the crash course

AI Agents for Marketing Crash Course

Learn how to deploy AI marketing agents across your go-to-market — the best tools, prompts, and workflows to turn your data into autonomous execution without writing code.

Download the PDF

Understanding the Limitations: ChatGPT vs. Traditional BI Tools

Using ChatGPT for data analysis is incredibly fast and intuitive, but it’s important to understand where it fits in your toolbox. It’s not a replacement for dedicated business intelligence platforms like Tableau, Power BI, or Looker Studio, but it serves a very different purpose.

Where ChatGPT Excels:

• Speed and Simplicity: There’s no faster way to go from a raw data file to a chart. You don't have to learn software, you just have to ask.
• Data Exploration: It's perfect for quick, exploratory analysis to find interesting patterns that you might want to dive deeper into later.
• Accessibility: It makes data analysis accessible to everyone, not just people with "analyst" in their job title.

Where It Has Limitations:

• It Creates Static Images: The chart ChatGPT gives you is a static .png file. It's not an interactive dashboard. You can’t hover over points to see data or filter results.
• It's Not Live Data: If your source data in the spreadsheet changes, your chart won't update. You have to re-upload the file and start a new chat from scratch every single time.
• Accuracy Checks are Manual: While powerful, AI can still occasionally misinterpret prompts or data structures. For reports that are guiding million-dollar decisions, you’ll always need a human to double-check the work. Also, be careful not to upload sensitive or private information to a third party.

Essentially, treat ChatGPT as your fast and agile data brainstorming partner. Use it to quickly analyze data exports, find stories in the numbers, and help you get an initial draft of visualizations you can later perfect or automate in other tools.

Final Thoughts

Creating a scatter plot with ChatGPT can turn a jumble of raw numbers into a clear visual insight in a matter of seconds. By preparing your data as a simple CSV file and using clear, conversational prompts, you can quickly analyze the relationships between key elements of your business, including things like your product's pricing versus sales numbers, or your website traffic against the number of weekly blog posts.

While ChatGPT is fantastic for creating stand-alone static visualizations one at a time, its real superpower shines when used as a launching point to get quick answers that can later be turned into fully fledged dashboards. We built Graphed to do exactly this - simply connect all of your company's data sources in one place to create dashboards, and let the dashboards work for you automatically, updating around the clock using real-time data - so that you can see accurate, up-to-date information for all elements of your business at a single glance, at any time of day.