How to Create a Pivot Table in Looker with AI

Creating a pivot table is one of the most effective ways to summarize large amounts of data, turning endless rows of information into a compact, organized summary. While tools like Looker (now part of Google Cloud) are powerful for this, the process has always required a manual, click-by-click approach. This article will walk you through the standard way to build a pivot table in Looker and then show you how modern AI-powered tools are changing the game entirely.

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 Is a Pivot Table and Why Use One in Looker?

At its core, a pivot table is a data summarization tool. It takes a detailed dataset - like a long list of individual product sales - and reorganizes it into a meaningful new table. You can "pivot" or arrange the data to highlight different relationships and insights.

Think of it this way: instead of a transaction log with 10,000 rows showing every single sale, you can create a pivot table to see:

• Total sales revenue for each product category, grouped by quarter.
• The average order value for customers from different countries.
• A count of deals closed by each sales rep in each region.

In Looker, pivot tables let you take the data from your company's database - which is modeled and organized by your data team - and slice it in ways that answer specific business questions without needing to write any SQL code yourself.

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

How to Create a Pivot Table in Looker: The Standard Method

Building a pivot table in Looker is a methodical process. It happens within Looker's "Explore" interface, which is the starting point for any analysis. Here’s a breakdown of the traditional steps involved.

Step 1: Get Started with an Explore

First, you need to choose an Explore. An Explore is a pre-built view of your data created by your data analysts, like "Sales Data," "Website Traffic," or "Customer Tickets." It serves as your canvas, giving you a curated set of data fields to work with.

Step 2: Select Your Dimensions and Measures

Next, you’ll see a list of available data fields on the left side of your screen. These are split into two essential categories:

• Dimensions: These are the qualitative, descriptive attributes you want to group your data by. Think of them as the labels for your rows and columns. Examples include Date, Country, Product Name, or Campaign Name.
• Measures: These are the quantitative, numerical values you want to calculate or aggregate. This is the data you're actually summarizing. Examples are Sum of Revenue, Average Session Duration, or Count of Orders.

To start, select at least one dimension and one measure. For instance, you could select "Region" (a dimension) and "Total Sales" (a measure) to see sales per region.

Step 3: Run the Query

After selecting your fields, click the "Run" button. Looker queries your database and displays the results in a simple data table. At this point, you'll have a flat table - for example, a list of regions in one column and their corresponding total sales in another.

Step 4: Pivot a Dimension

This is where the magic happens. To create a pivot table, you need to choose one of your dimensions to serve as the columns of your new table.

1. Find the dimension you want to pivot in the sidebar (the one you already selected).
2. Click the small gear icon next to its name.
3. Select the "Pivot" option from the dropdown menu.

When you select "Pivot" on a dimension (let's say you chose "Product Category"), Looker takes all the unique values from that dimension ("Shirts," "Pants," "Shoes") and turns them into individual columns.

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 5: Run the Query Again to See the Result

Click "Run" one more time. Looker will now display your data in a classic pivot table format. If you used "Region" for your rows, "Product Category" for your columns, and "Total Sales" as your value, you'll get a clean grid showing the total sales for each category within each region - a much faster way to spot trends than scrolling through a huge list.

From here, you can continue to add more dimensions, measures, apply filters, and customize the visualization to turn your pivot table into a bar chart, line graph, or other visuals to add to a dashboard.

The Challenges of Building Reports the Manual Way

While the step-by-step process in Looker is straightforward for those who know the tool, it highlights a few common bottlenecks that marketing and sales teams often face.

• A Steep Learning Curve: To get value from Looker, you have to learn its specific vocabulary and user interface - Explores, dimensions, measures, pivots, filters. This proficiency takes time and is often unrealistic for busy team members who just need a quick answer. Many companies find it takes dozens of hours for an employee to become truly confident navigating a traditional BI tool.
• Iterative and Time-Consuming: The process of exploring data involves a lot of clicking, running queries, waiting, adjusting, and running again. Answering a simple follow-up question ("Okay, but what does this look like for only mobile traffic?") means finding the right filter, applying it, and re-running the whole report. Each tweak interrupts your flow of analysis.
• Dependency on Data Teams: Everything you can do in Looker depends on how your data team has set up the underlying data models (known as LookML). If you need to analyze data that isn't already included in an Explore - or if you want to combine data from different sources like Salesforce and your Facebook Ads account - you usually have to submit a request and wait for a data analyst to make those changes.

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

The Rise of AI: From Building Reports to Asking for Them

This is where a fundamental shift is happening. Instead of you needing to learn the language of a data tool, AI-powered platforms learn the language of your data. This lets you skip the clicks and go straight to the conversation.

Consider the process again, this time with an AI data analyst:

The Looker (Manual) Process:

1. Find the right "Explore."
2. Scroll to find and click the "Region" dimension.
3. Scroll to find and click the "Product Category" dimension.
4. Scroll to find and click the "Total Sales" measure.
5. Click "Run."
6. Go back and find the "Product Category" dimension, click the gear icon, and select "Pivot."
7. Click "Run" again.
8. Analyze the resulting table.

The AI-Powered (Conversational) Process:

1. Ask, "Show me a pivot table of total sales by region and product category for Q4."
2. The tool builds it instantly.

This new, conversational approach removes nearly all the friction. Instead of navigating menus and remembering process steps, you just describe the exact final report you want to see.

Key Benefits of this AI-Driven Approach

• It's Instantaneous: You get from question to insight in seconds. Follow-up analysis is just as fast. You can ask "Okay, now show that as a bar chart" or "Which region grew the fastest compared to last quarter?" without starting from scratch.
• It's Accessible to Everyone: Your most junior team members can get answers without needing any technical knowledge or BI tool training. If they can describe what they need in an email, they can build a report. This creates a more data-driven culture because no one has to be a "data person" to get valuable information.
• It Unifies Your Data Sources: Most marketing and sales questions don't live in a single system. You need to connect ad spend from Facebook Ads to sales data in Shopify to lead data in HubSpot. AI platforms are built to connect these sources automatically, allowing you to ask cross-platform questions that are impossible to answer within a single tool's native reporting.

Final Thoughts

Looker is a powerhouse BI tool, and creating a pivot table is a foundational skill for anyone conducting an analysis. Understanding the manual process of selecting dimensions, choosing measures, and pivoting the data is valuable. However, the future of data analytics is one where this manual, click-based assembly is no longer required for most day-to-day business questions.

This is precisely why we built Graphed. We wanted to eliminate the steep learning curves and manual report-building that slow teams down. We connect directly to all your key data sources - like Google Analytics, Shopify, Salesforce, HubSpot, and your ad platforms - so all your data is in one place. From there, you can just ask questions in natural language, and our AI builds live, interactive dashboards for you in seconds. Instead of you learning a BI tool, our tool learns what you need.