How to Create a Metrics Dashboard in Looker
Creating a good metrics dashboard in Looker turns a sea of raw data into a clear story about your business performance. Instead of drowning in spreadsheets, you get a single, interactive view that shows you what’s working and what isn’t. This guide will walk you through the essential steps, from understanding Looker's core concepts to building and arranging your very own interactive dashboard.
First Things First: Getting Your Data Connected
Before you can build anything, Looker needs access to your data. Unlike tools that have you upload a CSV, Looker connects directly to your company’s SQL-compatible database (like BigQuery, Redshift, or Snowflake). This is a huge advantage because it means your dashboards are always pulling live, up-to-date information. You're never looking at a stale report from last week.
Setting up this initial connection is typically a job for a database administrator or a dedicated Looker admin. They work behind the scenes to establish a secure link between Looker and your data warehouse, ensuring all the right permissions are in place. Once that connection is made, you and your team can get to work building.
The Heart of Looker: A Quick Introduction to LookML
The most important thing to understand about Looker is its modeling layer, called LookML. This isn't something you can skip over, as it’s the foundation for every single chart and metric in your dashboard. Think of LookML as the dictionary or "single source of truth" for your company's data.
It’s a layer of code that sits between your database and your dashboards. In the LookML model, developers define all your business logic centrally. This prevents situations where one person calculates "revenue" one way, and another person calculates it slightly differently. In Looker, "revenue" is defined once in the LookML, so everyone in the company is working from the same playbook.
Defining Your Calculations (Measures)
In Looker, a quantitative metric you want to track - like a total, average, or count - is called a measure. Measures are the numbers you want to see on your dashboard. They aggregate your data.
For example, a LookML developer might define a measure for "Total Revenue" like this:
This simple block of code tells Looker: "To calculate total revenue, find the column named 'sale_price' and add up all its values." Complex calculations are possible, but the idea is the same: define the business logic once, and it’s available for everyone to use.
Defining Your Categories (Dimensions)
If measures are your numbers, dimensions are the categories you use to group or filter those numbers. Dimensions are how you slice and dice your data. Think of phrases like "by date," "by country," or "by product category."
A dimension for "Country" might be defined as:
Once these dimensions and measures are defined in your LookML model, they appear as simple point-and-click options for everyone building reports. This is what allows non-technical users to build sophisticated analyses without ever writing a line of SQL.
Building Your Visualizations: From Explore to Look
With your data connected and LookML model in place, you can start creating individual charts and data tables. In Looker terminology, a single saved visualization is called a Look. You create Looks in an interface called an Explore.
An Explore is a starting point for asking questions of your data. Think of it as a sandbox built around a specific business topic, like "Order Analysis" or "User Activity."
Step-by-Step: Creating a Look
Let's say you want to create a line chart showing daily revenue for the last month. Here’s how you'd do it:
1. Choose your Explore: Navigate from the sidebar and select the "Orders" Explore.
2. Select your fields: On the left side of the screen, you’ll see a list of available Dimensions and Measures. From the "Orders" section, you’d click a dimension like "Created Date." From the Measures section, click "Total Revenue."
3. Add a filter: You don’t want all revenue from all time. Find the "Filters" section, choose the "Created Date" dimension, and set the condition to "is in the last 30 days."
4. Run the query: Click the "Run" button. Looker instantly generates the SQL query in the background, sends it to your database, and gets the results back. You'll see a data table with dates and corresponding revenues.
5. Choose a visualization: Above the data table, click the "Visualization" tab. Looker offers many chart types. Select the "Line" chart option. Voila, you have a line chart!
6. Save your work: Click the gear icon in the top right and select "Save as a Look." Give it a clear name like "Daily Revenue - Last 30 Days" and save it to a folder.
Repeat this process for every key metric you want on your dashboard - "Users by Country," "Sales by Product Category," etc. Each one becomes a saved Look ready to be arranged on a dashboard.
Assembling Your Dashboard: Putting the Pieces Together
Now for the fun part: combining all your saved Looks into a comprehensive dashboard. A dashboard is simply a collection of these charts and tables (called "tiles") arranged on a single screen.
Step 1: Create a New Blank Dashboard
Start by navigating to the folder where you saved your Looks. In the top right corner, click the "New" button and select "Dashboard." Give it a name like "Executive Summary" or "Marketing Performance" and click "Create Dashboard." This will give you a blank canvas to work with.
Step 2: Add Tiles to Your Dashboard
Once you’re in your new blank dashboard, click "Edit Dashboard." You’ll now see an option to "Add Tile." A pop-up will appear, giving you two main choices:
Choose a Look: This is the most common option. It will let you search for and select the Looks you just created. Select "Daily Revenue - Last 30 Days" and it will appear on your dashboard as a tile.
From an Explore: Sometimes you need a simple number or a quick chart that you don't feel needs to be saved as its own Look. This option lets you build a visualization from scratch directly on the dashboard, following the same process as in an Explore. It’s great for one-off additions.
Add all the Looks you created one-by-one until your dashboard has all the components it needs.
Step 3: Arrange and Resize Tiles
Once your tiles are on the dashboard, you can drag, drop, and resize them to create a logical flow. The best dashboards tell a story. A common layout pattern is:
Top Row: High-level KPIs and single-value visualizations. The most important numbers (e.g., "Total Revenue This Month," "New Users This Week").
Middle Rows: Trends and breakdowns. Your line charts showing performance over time, and bar or pie charts breaking down metrics by category.
Bottom Rows: Granular data. Raw data tables for those who want to see the details behind the charts.
Play around with the layout until it’s easy to read and the relationships between the charts are clear.
Making Your Dashboard Interactive With Filters
A static dashboard is nice, but an interactive one is truly powerful. Looker lets you add dashboard-level filters that can control all (or some) of the tiles at once. This empowers users to answer their own follow-up questions without asking you to create a new report.
To add a filter, go into "Edit Mode" and click "Filters" in the top toolbar, then "Add Filter."
For example, you could add a "Date" filter. When you configure the filter, you’ll then go through a simple workflow to tell Looker which "date" field in each tile this dashboard filter should update. For your "Daily Revenue" tile, you link it to the orders.created_date field. Do this for all relevant tiles.
Now, when a user changes the dashboard's date filter from "Last 30 Days" to "Last 90 Days," every connected tile on the dashboard updates instantly. This turns a one-dimensional report into a dynamic analytical tool.
Final Thoughts
Building a dashboard in Looker is a methodical process that offers immense power and flexibility once you grasp the core concepts of LookML, Explores, and Looks. By defining your business logic centrally and building visualizations piece by piece, you can create a reliable and interactive single source of truth for your entire team.
While Looker provides incredible depth, getting started requires a significant investment in learning LookML and the manual click-by-click process of building each chart. For many marketing, sales, and e-commerce teams, that time could be better spent on strategy. At Graphed, we remove that friction completely. We connect your key data sources - like Google Analytics, Shopify, Facebook Ads, and Salesforce - and allow you to create powerful, real-time dashboards just by asking questions in plain English. Instead of learning a new BI platform, you can just ask, "Show me a dashboard comparing Facebook Ads spend vs. revenue by campaign for the last 30 days," and we build it for you instantly.