How to Make a Scatter Plot in Looker with AI
A scatter plot is one of the most effective ways to understand the relationship between two different variables in your data. It helps you answer critical questions like, "Does spending more on ads lead to more sales?" or "Are more sales calls connected to a higher close rate?" This post will walk you through how to create a scatter plot in Looker, covering both the traditional manual method and the faster, more intuitive way using AI.
What is a Scatter Plot, Anyway?
Before jumping into Looker, let's quickly review what a scatter plot does. Imagine you have a big table of data. A scatter plot takes two columns of numbers from that table and plots them on a graph. One number becomes the position on the horizontal line (the x-axis), and the other becomes the position on the vertical line (the y-axis). The result is a collection of dots, where each dot represents a single row from your data table.
By looking at the pattern of these dots, you can instantly see if there's a relationship between the two variables. This visual gut-check is incredibly valuable for spotting trends that you’d miss by just staring at a spreadsheet.
You can identify a few common patterns:
Positive Correlation: As one variable increases, the other variable also tends to increase. The dots will follow a pattern from the bottom-left to the top-right of the graph. Example: The more you spend on ads, the more revenue you generate.
Negative Correlation: As one variable increases, the other variable tends to decrease. The dots will form a pattern from the top-left to the bottom-right. Example: The higher the discount you offer, the lower your profit margin per item.
No Correlation: There's no obvious relationship between the two variables. The dots are scattered all over the graph with no clear direction. Example: The number of likes on your recent Instagram post has no connection to your company's support ticket volume.
Scatter plots are essential for marketers, sales leaders, and business owners who need to connect the dots between actions and outcomes.
Creating a Scatter Plot in Looker: The Manual Method
Looker is a powerful business intelligence tool, but its power comes with a certain level of complexity. If you're familiar with its interface, creating a scatter plot is straightforward, but it requires several precise steps. Let's walk through the manual process of creating a plot to see how ad spend correlates with revenue.
Step 1: Start with an Explore
Everything in Looker begins at the "Explore" level. An Explore is a curated starting point for a query, built from your underlying data model. You first need to navigate to the appropriate Explore that contains the data you want to analyze, like a "Marketing Performance" Explore that has both advertising and sales data in it.
Step 2: Select Your Dimensions and Measures
Once you're in an Explore, you'll see a list of available fields on the left side, organized into "Dimensions" and "Measures."
Dimensions are the descriptive, categorical fields that you group by, like Campaign Name, Product Title, or Country.
Measures are the numerical, aggregatable fields that you calculate, like Total Ad Spend, Total Revenue, or Count of Clicks.
For a scatter plot, you’ll typically need two measures (one for your x-axis and one for your y-axis) and one dimension to define what each individual dot on the plot represents. For our example, let's select:
Dimension: Campaign Name
Measure 1: Total Ad Spend
Measure 2: Total Revenue
Step 3: Run the Query and Get Your Data
After selecting your fields, click the green "Run" button. Looker will query your database and return a data table with three columns: Campaign Name, Total Ad Spend, and Total Revenue. This table forms the foundation for your visualization.
Step 4: Choose the Scatter Plot Visualization
Above the data table, you'll find the "Visualization" pane. Looker offers dozens of chart types. Click on the three dots to open the full list and select the "Scatterplot" option. Looker will make a first attempt at creating the chart, but you’ll likely need to fine-tune it to get what you want.
Step 5: Configure the Visualization Settings
This is where things can get a bit tricky and require some trial and error. Click the gear icon in the top right corner of the Visualization pane to open the settings menu. Here’s what you’ll need to adjust:
Plot Tab
Here, you tell Looker how to draw the chart. You need to assign your fields to the right axes:
X-Axis: Select your "Total Ad Spend" measure.
Y-Axis: Select your "Total Revenue" measure.
You may also configure things like the size of the dots (you could tie size to a third measure like Return on Ad Spend) and whether to group the data series.
X and Y Axis Tabs
These tabs control the appearance of your axes. You can:
Change the axis names (e.g., from "Total_Ad_Spend" to "Ad Spend ($)").
Adjust the scale (e.g., from linear to logarithmic, which is useful when your data has extreme outliers).
Show or hide gridlines for better readability.
After configuring everything, your scatter plot will correctly show the relationship between ad spend and revenue for each of your campaigns. While this process is effective, it depends on you knowing which Explore to use, how to distinguish between dimensions and measures, and how to navigate multiple settings menus. For non-technical users, this learning curve can be steep.
The Smarter Way: Using AI to Generate a Scatter Plot in Looker
Recent advancements in generative BI are fundamentally changing how users interact with platforms like Looker. Instead of clicking through menus and manually configuring charts, you can now use natural language - plain, conversational English - to ask for what you need.
This approach streamlines the entire workflow, making sophisticated data analysis accessible to everyone, not just data analysts.
Step 1: Open the Conversational AI Interface
Modern versions of Looker include a generative AI component, often accessible via a button that says "Start a conversation" or an input bar where you can ask a question about your data. This is your direct line to the AI data analyst built into the platform.
Step 2: Describe the Chart You Want in Plain English
Instead of manually selecting fields, just type what you want to see. The key is to be descriptive but conversational. For our same goal of analyzing ad spend vs. revenue, you could type a prompt like:
Create a scatter plot comparing total ad spend vs total revenue for each campaign.
Want something different? You can ask for anything:
"Show me a scatter plot of customer acquisition cost against customer lifetime value for new customers this year."
"Plot sales team call volume versus number of deals closed per day last quarter."
Step 3: Let the AI Generate the Visualization
As soon as you enter the prompt, the AI gets to work. It interprets your request, identifies the correct fields in the data model (like Total Ad Spend and Total Revenue), figures out that Campaign Name is the right dimension, runs the query, selects the scatter plot chart type, and automatically configures the x and y axes for you. In a matter of seconds, the exact chart you requested appears, fully configured.
Step 4: Refine Your Chart with Follow-up Questions
Here’s where conversational AI truly shines. The initial chart often sparks new questions. Instead of going back into the settings menus to make adjustments, you can continue the conversation with the AI. For example:
You see an outlier and want to exclude it: "Great, now filter out campaigns that spent more than $10,000."
You want to segment the data differently: "Update the chart and color the dots by the marketing channel."
You want to get a better sense of the trend: "Could you add a linear trend line to this plot?"
Each time, the AI understands your follow-up request in context and instantly updates the visualization. The process feels less like configuring software and more like brainstorming with a data analyst who does all the tedious work for you.
Why Conversational AI is a Game-Changer for Looker Users
The shift from manual configuration to conversational prompting represents a massive leap forward in making data accessible.
It lowers the barrier to entry. You no longer need to complete an extensive training course to become proficient with Looker. If you know what question you want to ask, you can get an answer. This empowers your whole team, from junior marketers to senior executives, to make better, data-driven decisions on their own.
It saves an incredible amount of time. A task that used to take several minutes of clicking, dragging, and configuring can now be done in seconds with a single sentence. This allows you to spend your time analyzing insights, not wrestling with a tool.
It fosters deeper data exploration. Because it's so easy to ask follow-up questions, users are more likely to dig deeper into the data. The "what if" scenarios that once seemed too time-consuming to explore are now frictionless, leading to more "aha" moments and better insights about what’s truly driving your business.
Final Thoughts
Scatter plots provide an incredibly clear view of the relationships hidden in your business data. While you can certainly build them manually in Looker, using conversational AI drastically simplifies the process, turning a technical task into a simple conversation and putting powerful analysis at your fingertips.
For many teams, however, the primary challenge isn't just creating a single chart but connecting all their scattered data sources in one place to begin with. We built Graphed to solve this exact problem. It allows you to connect your marketing and sales tools - like Google Analytics, Shopify, Facebook Ads, and Salesforce - in just a few clicks. From there, you can ask for a scatter plot or build an entire real-time dashboard just by describing what you want in plain English. You get all the power of an enterprise BI tool, with none of the complexity.