How to Create a Real-Time Dashboard with Looker

A real-time dashboard gives you a live pulse on your business, helping you move from gut feelings to data-backed decisions. One of the most powerful tools for this is Looker. This article will guide you through the process of creating a real-time analytics dashboard in Looker, covering the essential groundwork, the step-by-step process of building your first visuals, and tips for making it genuinely useful.

Understanding “Real-Time” Data in Looker

Before you build, it’s important to understand what "real-time" means in the context of a business intelligence tool like Looker. It doesn’t mean your dashboard updates every millisecond like a stock ticker. Instead, Looker’s real-time capability comes from its architecture: it queries your live database directly.

Unlike some tools that require you to extract and load data in batches, Looker talks to your data warehouse (like Google BigQuery, Amazon Redshift, or Snowflake) the moment you load a dashboard. This means you’re seeing the most up-to-date information available in your database. This approach has a few key benefits:

• Freshness: The data is as fresh as the information in your warehouse, not from a static CSV you downloaded last Monday.
• Consistency: Everyone in the company queries the same live source, leading to a single source of truth.
• Speed: While dependent on your database's performance, it eliminates manual data export and upload steps that can slow down your reporting cycle.

Looker uses caching to speed up performance for frequently viewed dashboards, but you always have the option to force a refresh to pull the absolute latest data from your warehouse.

Before You Build: The Essential Groundwork

Jumping straight into building a dashboard without setting the foundation is a recipe for frustration. A little preparation goes a long way and ensures your dashboard is both accurate and useful.

Step 1: Connect Your Data Source

Looker doesn’t store your data, it’s a window into the data you already have. This means the very first step - typically done by a data administrator - is to establish a connection between Looker and your SQL database. This is the pipeline through which all your data will flow.

For most business users, you won't be setting up this connection yourself, but it's important to know it exists. Your company's data team will have already configured Looker to securely talk to your primary data sources, making them available for you to analyze.

Step 2: Get Familiar With LookML

Looker's biggest differentiator - and its main learning curve - is LookML. LookML is a modeling language that works as a translation layer between your database and the charts you build. Your data team uses it to define all of your company's business logic in one place.

Think of it like this: your database is a giant kitchen full of raw ingredients (columns like user_id, created_at, order_value). A business user doesn’t want to mess with raw ingredients, they want a menu. LookML is that menu. It pre-packages your raw data into understandable dishes:

• Dimensions: Segments you can group by (e.g., Country, Traffic Source, Product Category).
• Measures: Calculations and aggregates you want to track (e.g., Total Revenue, Average Order Value, Customer Count).

By defining "Total Revenue" once in the LookML model, everyone in the company gets the exact same number every time. It prevents a scenario where the marketing team calculates revenue one way, and the finance team calculates it another.

You probably won't be writing LookML yourself, but you will interact with its output: Explores. An Explore is a pre-built view of your data that serves as a logical starting point for asking questions.

Building Your Real-Time Dashboard: A Step-by-Step Guide

Once your groundwork is done, you're ready to start building. The workflow generally moves from asking a single question to combining multiple answers into a comprehensive dashboard.

Step 1: Start from an “Explore”

An Explore is where your analysis begins. Your data team will have created various Explores around core business concepts, such as “Users,” “Orders,” “Sessions,” or “Tickets.” Your first step is to choose the one that relates to the question you want to answer.

For example, if you want to understand website traffic trends, you’d likely start with the “Sessions” Explore. If you want to analyze sales data, you’d use the “Orders” Explore.

Step 2: Build Your First Visualization (a "Look")

Once you’re in an Explore, you'll see a user-friendly interface divided into two main parts: Dimensions and Measures. This is where you compose your query without writing any SQL code.

Let's walk through an example. Goal: Create a chart showing weekly revenue from your top 5 US states over the last 90 days.

1. Select Dimensions: You are grouping by time and location, so from your list of dimensions, you'd select Order Date (and specify "Week" as the timeframe) and State.
2. Select a Measure: You want to know the revenue, so from your list of measures, you’d select Total Revenue.
3. Add Filters: You're only interested in the US for a specific time period. You’d add two filters:
4. Set a Row Limit: Since you only want the top 5, you'd set the row limit to 5 and sort descending by Total Revenue.

Click "Run." Looker translates your clicks into a SQL query, sends it to your live database, and returns the results in a data table. From there, you can use the "Visualization" tab to turn that table into a bar chart, line chart, map, or another format. Once you're happy with it, click "Save" and name it. In Looker, a saved visualization is called a Look.

Step 3: Assemble Your "Looks" into a Dashboard

A dashboard is a collection of Looks organized on a single page. After creating a few individual Looks that answer related questions (e.g., one for revenue, one for top products, one for user location), you can combine them to tell a bigger story.

• Create a new Dashboard and give it a name (e.g., “Q3 Marketing Performance”).
• Add your saved Looks to the dashboard. You can add them one by one.
• Arrange the charts (called "tiles") on the grid by dragging, dropping, and resizing them. A good layout guides the viewer’s eye from high-level summaries at the top to more detailed breakdowns at the bottom.

Step 4: Make Your Dashboard Interactive with Filters

Here’s what makes a dashboard truly powerful. Instead of having hard-coded filters in each Look, you can add dashboard-level filters that apply to multiple charts at once. A common example is a date range filter.

By adding a "Date" filter to your dashboard, a user can instantly change the timeframe for all tiles — from "last 7 days" to "this quarter" — without having to edit each chart individually. This transforms your dashboard from a static report into an interactive analytical tool where your team can explore the data on their own.

Keeping Your Dashboard Data Fresh

Your dashboard is live, but how "live" is it? As mentioned, Looker uses a caching system to deliver results quickly. When a user runs a query, Looker stores the result for a set period. If another user runs the exact same query within that time, Looker serves the cached result instantly instead of querying the database again.

While this is great for performance, sometimes you need the absolute latest data. On any dashboard or Look, you can click the three-dot menu and select "Clear cache & refresh." This forces Looker to ignore the saved results and run new queries against your live database.

Admins can also configure dashboards to automatically refresh on a schedule, such as every 15 minutes or every hour. This is useful for high-priority operational dashboards viewed by many people, but be mindful that frequent refreshing can put a strain on your database and may incur costs.

Tips for an Effective Looker Dashboard

Building a dashboard is one thing, building a good dashboard is another. Here are a few best practices to keep in mind:

• Start With a Key Question: A good dashboard answers a specific business question, like "How are my paid campaigns performing this month?" or "What is our sales team's current pipeline velocity?" Don't just show data for the sake of it.
• Tell a Story: Organize your charts logically. Start with high-level KPIs (Key Performance Indicators) at the top, and then provide more detailed breakdowns below. A user should be able to grasp the main takeaways at a glance.
• Keep it Clean and Simple: Overloading a dashboard with too many charts, colors, or metrics makes it unusable. Focus on the few metrics that matter most. An effective dashboard is one that is easily understood.
• Use Descriptive Titles: Don't leave your charts with generic titles. Change "Count of Sessions" to "Weekly Website Sessions - Last 90 Days." Be specific so anyone can understand the chart's purpose without having to inspect it closely.

Final Thoughts

Creating a real-time dashboard in Looker is a methodical process. It involves a strong data foundation built on LookML, using "Explores" to build individual visualizations called "Looks," and then composing those into interactive dashboards that query your live database directly. While it's an incredibly powerful platform for enterprise business intelligence, the workflow requires technical setup, a well-defined data model, and user training to master.

That structured process is great for large organizations, but we've found that many marketing and sales teams need a much faster path from data to dashboard. For that reason, we built Graphed . We connect directly to your most common marketing and sales tools — Google Analytics, Shopify, Facebook Ads, Salesforce — and let you create real-time dashboards using simple, natural language. Instead of learning new software or waiting on a data team, you can just ask, "Show me a comparison of last month's ad spend vs. revenue by campaign," and watch as a live dashboard is built for you in seconds.