How to Create a Company Dashboard in Looker with AI
Creating a company-wide dashboard in Looker is a fantastic way to get a single, reliable view of your business performance. But let's be honest: the traditional process can be slow and often requires help from a data analyst. This guide will show you how to leverage Looker's new AI features to build a powerful company dashboard faster and more intuitively, even if you don't live inside Looker all day.
What is a Company Dashboard Anyway?
A company dashboard is a centralized, visual report that shows your business's most important key performance indicators (KPIs) in one place. Instead of logging into Salesforce for sales data, Google Analytics for website traffic, and QuickBooks for financial health, a company dashboard pulls it all together. Its purpose is to give anyone, from a founder to a department head, a quick, accurate snapshot of business health at a glance.
A typical company dashboard might track metrics from across the business, including:
Sales KPIs: New MRR (Monthly Recurring Revenue), Deals Closed, Sales Pipeline Value, Win Rate.
Marketing KPIs: Website Sessions, Leads Generated, Cost Per Acquisition (CPA), Campaign ROI.
Financial KPIs: Revenue vs. Expenses, Gross Margin, Customer Lifetime Value (LTV).
Product KPIs: Daily Active Users, Feature Adoption Rates, Churn Rate.
Putting these on a single screen helps you spot trends, understand how different departments impact each other, and make better-informed decisions without digging through a dozen different apps and spreadsheets.
Why Looker is a Go-To for Company Dashboards
While there are many BI tools out there, Looker (now part of the Google Cloud family and often referred to within Looker Studio) has become a favorite for data-driven companies for a few key reasons. It’s built around a powerful data modeling layer called LookML. Think of LookML as a set of instructions you write one time to teach Looker your business logic. You define what a "customer" is, how to calculate "revenue," and where to find "active users."
Once this foundation is set, everyone in the company can explore data using the same definitions, creating a true "single source of truth." This prevents the common scenario where the marketing team and the sales team pull slightly different numbers for the same metric, causing chaos.
The downside? Historically, that power came with a steep learning curve. The person building the dashboards needed to be fairly technical.
The Old Way vs. The New Way: AI Changes Everything
The traditional path to a Looker dashboard usually looked something like this:
A business leader needs a new report.
They file a ticket or send an email to the data team.
An analyst or data engineer has to find time to build the visualizations.
After a few days (or weeks), the dashboard is ready.
If you need a change? Back to step one.
This process creates bottlenecks and slows down decision-making. Marketers, sales leaders, and founders need to be able to explore data and get answers themselves without waiting in line. This is where AI comes in.
Google has integrated its powerful AI, Gemini, directly into the Looker experience. Now, instead of dragging, dropping, and configuring fields manually, you can simply describe what you want to see in plain English. This dramatically lowers the barrier to entry, empowering more people to get hands-on with their data.
How to Build Your Company Dashboard with Looker AI
Let's walk through the steps to create a company dashboard using Looker's AI-powered natural language features. This assumes your data sources are already connected to Looker and your data team has set up the foundational LookML models.
Step 1: Start with The "Explore" Interface
Looker's "Explore" is the sandbox where you build reports and charts. You'll choose a starting point, which is typically a pre-built dataset curated by your data team - like "Sales Data" or "Website-Traffic." Once you're in an Explore, instead of seeing the familiar sidebar with dimensions and measures, you’ll focus on the magic chat box at the top powered by Duet AI.
Step 2: Use Natural Language to Create Your First Visualization
This is where the fun begins. Think about the first metric you want on your dashboard. Let's say it's sales revenue over time. In the prompt box, you can simply type:
Show me total weekly sales revenue for the past 6 months as a line chart.
Looker's AI will interpret your request, select the correct data fields from the underlying LookML model, and generate the line chart for you. You didn't have to find the "Sales Date" field, the "Revenue" measure, or know how to pivot the data by week. You just asked for what you wanted.
Let’s try another one, this time for marketing:
What are our top 5 traffic sources by user count this quarter? show it as a bar chart.
Again, the AI gets to work and creates a clean bar chart showing your traffic breakdown. You can continue this process for every KPI you want to track, from financial metrics to product engagement.
Step 3: Refine and Ask Follow-up Questions
Often, an initial chart sparks more questions. The AI in Looker is conversational, meaning you can ask follow-up questions to drill down further. After generating your sales chart, you might ask:
Now, break that down by sales team.
Looker will update the visualization, perhaps turning the single line chart into a multi-line chart or creating a stacked bar chart. This iterative process of questioning and refining is incredibly powerful because it mirrors how we naturally think about data. You find an insight, which leads to another question, and another, until you get to the root of the issue.
Step 4: Add New Calculations on the Fly
What if you need a metric that doesn’t exist yet, like a conversion rate? In the past, you might have had to go back to the data team to request they add a new calculation to the LookML model. With the AI integration, you can often create it yourself.
For example, if you have session and conversion data, you could ask:
Show me the conversion rate, calculated as total conversions divided by total sessions, grouped by month.
The AI will understand the required logic, perform the calculation, and visualize the result, saving you a massive amount of time.
Step 5: Assemble and Save Your Dashboard
Once you’ve created a chart you're happy with, find the option to "Save" or "Add to Dashboard." You can create a brand new dashboard (let's call it "Company Health Overview") or add your chart to an existing one.
Repeat this process for all your key metrics. Create a chart for website traffic, another for new leads, one for pipeline value, and one for active users. As you add each visualization (or "Tile," in Looker terminology) to your new dashboard, you can drag and drop them to arrange your preferred layout. You can also add text tiles to give context, name each section, and apply dashboard-level filters, like a date range selector that updates all the charts at once.
Tips for Better Looker AI Prompts
Just like with tools like ChatGPT, the quality of your output depends on the quality of your input. Here are a few tips to get better results from Looker’s AI:
Be Specific: Don't just say "show me traffic." Instead, try "Show me new users from organic search weekly over the last 90 days." The more detail, the better.
Request a Chart Type: If you have a preference, specify it. Add phrases like "...as a bar chart," "...create a table," or "...show this as a pie chart" to the end of your prompt.
Use Your Business Vocabulary: Since the AI is powered by the LookML models, try using the terms your business uses. If your company calls leads "MQLs," use that in your prompt.
Start Simple and Iterate: Begin with a broad question and then use follow-up prompts to narrow it down. This is often easier than trying to write one perfect, complex prompt from the start.
Final Thoughts
Building a comprehensive company dashboard no longer has to be a complex project that only your data team can handle. Looker’s new AI capabilities make data exploration more accessible and conversational, allowing anyone to build the reports they need by simply asking questions in plain English.
While Looker has made huge strides, we've focused on taking this simplicity even further. With Graphed , you get the power of a natural language interface without having to manage a complex data modeling layer like LookML. We offer one-click integrations with your data sources like Google Analytics, Shopify, Facebook Ads, and Salesforce. Just connect your accounts, and our AI learns the data structure for you, so you can start building real-time dashboards immediately, simply by describing what you want to see.