How to Do Data Analysis in Looker
Looker is a business intelligence powerhouse for drilling into your company's data. If you’re just getting started, it can feel a little intimidating, but its real strength lies in an intuitive interface that lets anyone ask complex questions without writing a single line of SQL. This guide will walk you through the core steps of performing data analysis in Looker, from running your first query to creating a shareable visualization.
First, Understand the Looker Environment
When you first log in to Looker, you're looking at a well-organized workspace designed to help you find, explore, and visualize data. While there are many features, you'll mainly interact with a few key areas to get started with your analysis.
Folders: This is where all saved content lives. Folders - which can be personal, shared, or for specific teams - contain Dashboards and Looks. Think of it as a file directory for all your reports.
Dashboards: Dashboards are collections of saved charts, graphs, and KPIs on a single page. They provide a high-level overview of business performance and are updated automatically with the latest data.
Looks: A "Look" is a single saved report. It’s essentially a visualization or data table that answers one specific question (e.g., "What was our revenue by month for Q3?"). Looks are the building blocks of Dashboards.
Explores: This is the most important part of Looker for data analysis. An "Explore" is your starting point for asking new questions. It's an interactive environment where you can select data points, apply filters, and build queries from scratch. We’ll be spending most of our time here.
The Heart of Looker Analysis: The 'Explore' Interface
To analyze data in Looker, you start with a powerful feature called an "Explore." It’s a curated and organized dataset that your data team has prepared for you. Instead of working with raw database tables, which can be messy and confusing, an Explore presents the data in a clean, logical format that's easy to understand and use.
Inside an Explore, all the available data fields are divided into two categories:
Dimensions
Dimensions are the "group-by" fields. They represent the different attributes of your data - the who, what, where, and when. You can think of them as the columns you would see in a spreadsheet that categorize your data. Common examples of dimensions include:
Date (e.g., Order Date, Sign Up Date)
User Type (e.g., New vs. Returning)
Geography (e.g., Country, State, City)
Product Category
Campaign Name
Customer ID
Dimensions are typically color-coded blue in the Looker interface.
Measures
Measures are the quantitative values - the numbers you can do math on. They represent things you want to count, sum, average, or calculate. Measures aggregate data across your dimensions. For example, if you group by the Country dimension, a measure like Total Sales will show the sum of all sales for each country. Common examples include:
Total Revenue
Count of Users
Average Order Value
Number of Sessions
Minimum, Maximum, or Average Purchase Price
Measures are typically color-coded orange in the Looker interface.
Understanding the difference between dimensions and measures is the fundamental first step. You'll select dimensions to group and segment your data, and you'll select measures to quantify that data.
Your First Data Analysis: A Step-by-Step Guide
Let’s walk through a practical example. Imagine we work for an e-commerce company and want to answer the question: "What were our total sales for each product category in the United States over the last 90 days?"
This is a classic business question that's easy to answer using Looker's Explore interface.
Step 1: Choose Your Explore
First, navigate to the "Explore" section from the main navigation menu. You will see a list of available Explores created by your data team. These might be named things like "Orders," "Users," "Website Traffic," or "Products."
For our question, we're interested in sales and product information, so the "Orders" or "Sales" Explore seems like the right place to start. Let's click on it.
Step 2: Select Your Dimensions and Measures
Once you’re in the Explore, you’ll see a left-hand sidebar containing all available dimensions and measures, neatly organized into groups.
To answer our question, here’s what we need:
A dimension for product category: Let’s find a field named "Product Category" in a group like "Products" and click on it.
A measure for total sales: Now, we need the number. Let’s look for a field named "Total Revenue" or "Total Sale Amount" and click on that.
As you click on each field, Looker adds it to your query builder at the top of the page. You now have a simple query ready to go: it will show you the total revenue for every product category in your database.
Step 3: Add Filters to Narrow Your Data
Our current query will show us all revenue for all time across all locations. We need to be more specific. Here's where filters come in.
The "Filters" section is just above the data display area. We need to add two filters based on our original question:
Filter by Date:
Find a date dimension, like "Order Date," in the sidebar and hover over it. You'll see an icon that looks like a funnel – click "Filter."
Alternatively, just click on the "Order Date" field. It will be added to the query, now click on the new "Order Date" column funnel button to filter by it.
This adds the field to the "Filters" section. Here, you'll get a dropdown with different criteria. Choose a time-based filter like "is in the last" and type "90 days."
Filter by Country:
Find a location dimension, like "Ship to Country" or "User Country," and click the filter icon.
This adds it to your filters. For the condition, select "is equal to" and type "United States" into the value box.
Filters are incredibly powerful. You can create complex conditions, like filtering for sales greater than $100, including users from specific traffic sources, or excluding a certain product type.
Step 4: Run the Query
With your dimensions, measures, and filters in place, simply click the Run button in the top right corner. Looker translates your selections into SQL, sends the query to your database, and returns the results in seconds.
You'll now see a data table with two columns: Product Category and Total Revenue, showing only data from the last 90 days in the United States.
Step 5: Visualize Your Results
A table of data is good, but a chart is often better for quickly understanding patterns. Looker makes it easy to switch between visualizations.
Above the data table, you’ll see a "Visualization" tab. Click it.
You'll see a series of chart icons: Bar, Column, Line, Area, Pie, Scatter Plot, etc.
For comparing sales across different categories, a column chart or bar chart is perfect. Click the icon, and Looker will instantly convert your data into a visual format.
You can customize the visualization by clicking the "Edit" gear icon. This lets you change colors, adjust axis labels, display values, and more to make your chart clear and easy to read.
Diving Deeper: Intermediate Analysis Techniques
Once you're comfortable with the basics, you can explore some of Looker's more advanced features without needing any technical help.
Pivoting Data
Pivoting adds another layer of grouping to your query, turning a dimension's value into new columns much like a pivot table in Excel or Google Sheets. This is useful for comparing trends across two dimensions.
Example: In our query above, what if we wanted to see monthly sales for each product category?
Add a date dimension to your query, such as "Order Month."
Instead of just adding a column, to pivot it just click on the pivoting icon that is shown when hovering over a dimension to pivot the column.
Rerun the query. Now, instead of one "Total Revenue" column, you’ll see columns for each of the last three months, showing revenue for each product category broken down by month. This is perfect for spotting trends.
Creating On-the-Fly Table Calculations
Sometimes you need to create a metric that doesn’t already exist as a measure. "Table Calculations" let you perform Excel-like formulas directly on your data results without needing to change the underlying code.
Example: After running our initial query, let's say we want to see what percentage of total revenue each product category contributes to.
In the Data section, on the far right, there's a button called "Add calculation."
A popup will appear. In the formula box, you can create a calculation using your existing fields. To calculate the percent of total, you would type a formula like this:
${total_revenue} / sum(${total_revenue})
Name your calculation (e.g., "Percent of Total"), choose a format (like Percent), and hit "Save". A new column instantly appears in your table with the calculated values.
Saving and Sharing Your Findings
Once you’ve uncovered an interesting insight, your analysis is easily shareable.
Save as a Look: Click the gear icon in the top right and select "Save as a Look." Give it a name, a description, and choose a folder to save it in. Now you can easily come back to this exact report at any time.
Add to a Dashboard: From the same menu, you can add your report to a new or existing dashboard. This lets you build a comprehensive overview by combining multiple related reports.
Download or Share: You can also download your data (as a CSV, Excel file, etc.) or the visualization itself (as a PNG or JPEG). There are also options to share a public link or schedule the report to be emailed to stakeholders on a recurring basis (e.g., every Monday morning).
Final Thoughts
Looker empowers team members to move beyond static reports and explore data freely, turning curiosity into valuable business insights. By mastering the core concepts of Explores, dimensions, measures, and filters, you can start answering your own questions and contribute to a more data-informed culture without needing to be a data specialist.
While Looker is a phenomenal tool for teams with dedicated data support, we know the initial setup and learning curve can be steep for businesses without those resources. Learning LookML models, Explores, and BI fundamentals still presents a hurdle. That's why we created Graphed. It lets you connect your data sources in seconds and ask questions using simple, natural language. Instead of clicking through menus to build queries, just describe what you want - like "Show me a dashboard of sales by product category vs Facebook Ad spend for the last quarter" - and instantly get a real-time dashboard of your data, making deep data analysis truly accessible to everyone on your team.