How to Create a Product Management Dashboard in Looker with AI

Building a great product requires more than just intuition, it demands a clear, data-driven view of how users are actually engaging with what you've built. A product management dashboard is your mission control for this, but creating one in a powerful tool like Looker can often feel like a task reserved for data analysts. This guide walks you through how to concept, plan, and build a powerful product dashboard in Looker, and how new AI capabilities are making this process faster and more accessible than ever before.

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

Start With Your "Why": Planning a Dashboard That Delivers Insight

Jumping straight into Looker to build charts without a plan is a recipe for a cluttered, unhelpful dashboard. The best dashboards are carefully planned to answer specific, critical questions about your product's health and user experience. Before you touch a single visualization, take a step back and define your objectives.

Step 1: Identify Your Core Questions

Get your team together and brainstorm the most important questions you need to answer. Move beyond vanity metrics and focus on what truly drives product success. Your questions might include:

• Are new users successfully completing our onboarding process?
• Which features are our most engaged users adopting?
• How has the recent feature launch impacted user retention?
• Where are users dropping off in the sign-up funnel?
• Is product usage growing, stagnating, or declining over time?

Each question will directly map to a metric and a visualization on your dashboard, ensuring every element has a clear purpose.

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

Step 2: Choose the Right Product Metrics

With your questions defined, you can now select the key performance indicators (KPIs) that will answer them. A popular framework for SaaS product management is Dave McClure's AARRR, also known as "Pirate Metrics."

Let's break down which metrics you might track for each stage:

Acquisition: How are users finding us?

While often seen as a marketing area, product managers should have a pulse on acquisition channels to understand user intent.

• Sign-ups by Source: Understand which channels (organic, paid, referral) bring in the most users.

Activation: Are users having a good first experience?

This is where you measure the "a-ha!" moment - the point where a new user finds value in your product.

• Activation Rate: The percentage of new users who complete a key action (e.g., creating their first project, inviting a teammate).
• Time to Value (TTV): How long it takes a new user to reach that activation event.
• Onboarding Funnel Completion: Track the drop-off rates at each step of your introductory flow.

Retention: Are users coming back?

Growth is meaningless if users don't stick around. Retention is the single best indicator of product-market fit.

• User Retention Rate / Churn Rate: The percentage of users who remain active (or become inactive) over a specific period.
• Daily/Weekly/Monthly Active Users (DAU/WAU/MAU): Tracks the overall size of your active user base. The ratio (e.g., DAU/MAU) can indicate stickiness.
• Session Frequency: How often are cohorts of users logging in?

Referral: Are users telling others?

This tracks your product's virality and word-of-mouth growth.

• Net Promoter Score (NPS): A direct measure of user satisfaction and willingness to recommend.
• Referral Invitations Sent: The number of users utilizing your referral features.

Revenue: Are we making money?

Finally, connect product usage to business outcomes.

• Customer Lifetime Value (CLV): The total revenue you can expect from a single customer.
• Average Revenue Per User (ARPU): Helps you understand the value of different user segments.
• Free-to-Paid Conversion Rate: For freemium models, this is the most critical revenue metric.

Getting Your Data into Looker

Looker works by sitting on top of a company's database (like Google BigQuery, Snowflake, or Amazon Redshift). Raw event data from your product - clicks, pageviews, sign-ups - is typically sent to this database using a tool like Segment or loaded directly from your application's backend.

The Role of LookML

You can't build meaningful reports without a bit of setup. This is where Looker's modeling language, LookML, comes in. Think of LookML as a "dictionary" for your data. A data analyst or engineer on your team will use it to define your business logic in one central place.

For example, instead of every product manager having to remember the complex SQL query to calculate "Weekly Active Users," an analyst defines it once in LookML. Now, "Weekly Active Users" is just a field you can drag and drop into any report, ensuring consistency for everyone.

This pre-defined layer is what makes exploring data in Looker powerful, but it's also what traditionally created a dependency on the data team to set everything up just right.

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

Building Your Dashboard: The Manual Way vs. The AI-Powered Way

Once your data is connected and your LookML model is in place, you can finally start building your dashboard visualizations.

The Traditional Method in Looker

Traditionally, building a dashboard in Looker meant creating each visualization one by one. The process generally looked like this:

1. Navigate to an "Explore," which is a starting point for querying a specific dataset (e.g., a "Users" Explore or a "User Events" Explore).
2. Select the dimensions (fields to group by, like "Sign-up Date" or "Device Type") and measures (calculations to perform, like "Number of Users" or "Average Session Length") you need.
3. Apply any necessary filters (e.g., date range, user segment).
4. Run the query to see the data table.
5. Choose a visualization type (like a line chart or a bar chart) and customize its appearance.
6. Save this visualization as a "Look."
7. Go to your dashboard, click "Add Tile," and select the Look you just created.
8. Repeat this process for every single chart on your dashboard.

As you can see, this process is powerful and highly customizable, but it's also time-consuming. It requires you to know exactly which dimensions and measures to select from potentially hundreds of options, and a single dashboard with 10-15 charts could easily take a few hours to assemble.

The New Looker AI Assistant

This is where things get exciting. Looker and many other BI platforms are integrating generative AI to streamline this entire workflow. Instead of manually clicking, dragging, and dropping fields, you can simply describe what you want to see in plain English.

Within Looker Studio's AI features, you can interact with a chat interface to build your analysis without manually finding every field. This conversational approach dramatically lowers the technical barrier and speeds up the entire process.

Consider these examples:

Simple Request:

• Your prompt: "Show me a line chart of Weekly Active Users for the last 12 months."

The AI assistant understands "Weekly Active Users" (thanks to the LookML definition), selects the right date fields, chooses a line chart visualization, and generates the chart for you instantly.

More Complex Request:

• Your prompt: "What are our top 10 most used features this quarter? Show it as a bar chart, and segment the users by their pricing plan."

Here, the AI parses the date range ("this quarter"), identifies the segmentation need ("pricing plan"), and generates a complex, stacked bar chart without you needing to find all those fields yourself.

Exploratory Request:

• Your prompt: "Compare the 7-day retention rate for users who signed up in January vs. users who signed up in April."

This kind of prompt allows you to go from a question to an insight in seconds. The AI creates the cohort analysis for you, letting you drill down follow-up questions like, "What was different about the April cohort's behavior in their first week?" This turns dashboard creation from a static, pre-planned activity into a dynamic conversation with your data.

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

Final Thoughts

Building an effective product management dashboard in Looker is about starting with clear questions, defining the right KPIs, and then translating those into clear, actionable visualizations. This process, once a painstaking, manual effort, is being transformed by AI, empowering product teams to self-serve insights and find answers in seconds instead of hours.

While Looker's AI is powerful, many teams find their data scattered across platforms that aren't easily piped into a centralized BI tool like Looker in the first place. This is where we designed Graphed to simplify the process. We allow you to connect all your sources in just a few clicks - from product analytics tools to CRMs like Salesforce and ad platforms like Google Ads - and then build dashboards using the same conversational AI approach. Our goal is to unify your data and give you instant answers, so you spend less time wrangling data and more time building better products.