How to Create a Trend Report in Looker with AI

Building a trend report often feels like a time-consuming chore, but spotting patterns in your data is one of the most powerful things you can do for your business. It tells you what's working, what's about to break, and where your next big opportunity might be. This guide will show you how to use Looker's built-in and AI-powered features to create insightful trend reports faster 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

What Exactly is a Trend Report and Why Bother?

At its core, a trend report is simply a visualization of how a particular metric changes over time. It helps you move beyond static numbers and see the story your performance data is telling. Instead of just knowing you made 1,000 sales last month, you can see if those sales are on an upward curve, flatlining, or declining after a recent peak.

Tracking trends is vital because it helps you:

• Be Proactive, Not Reactive: See a dip in website traffic before it affects your lead generation. Notice a product category gaining momentum and ramp up marketing before your competitors do.
• Understand Cause and Effect: Did that email campaign actually cause a spike in sign-ups? By layering events over your trendline, you can visually connect your actions to their results.
• Forecast Future Performance: By understanding historical patterns, you can make smarter predictions about future revenue, user growth, or inventory needs.

For example, an e-commerce store owner might create a trend report on sales for a specific product. If they see a steady increase over the past six months, they can confidently order more inventory. But if they see a sudden, sharp decline, it's a signal to investigate, perhaps by checking customer reviews or competitor activity.

Start with a Strong Foundation: Preparing Your Data

Before you get to the exciting AI features, your trend report is only as good as the data powering it. Sloppy data leads to misleading trends. Getting this part right in Looker is about ensuring your LookML model - the code that defines your data's dimensions and measures - is in good shape.

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

Clean Up Your LookML Model

Think of your LookML model as the blueprint for your data. A strong blueprint makes building anything easy. Work with your data team to ensure:

• Timeframes are Clearly Defined: Your model should have clean, accessible date dimensions (e.g., created_date, order_date, signup_date). This is the backbone of any trend analysis.
• Metrics are Accurate: Your key measures (like total_sales, user_count, or ad_spend) should be accurately defined and calculated consistently.
• Data granularity is understood: Is your data recorded daily, hourly, or weekly? Knowing this helps you choose the right timeframe for your report and avoid misleading interpretations. Daily data is often best for spotting granular trends.

Your goal is to have a single source of truth that powers all your reporting. When everyone in the company builds reports from the same clean LookML model, you avoid conflicting numbers and confusing conclusions.

The First Step into AI: Using Looker’s Built-in Forecasting

Looker has included simple, AI-powered forecasting for a while now, and it's the easiest way to start looking into the future. A forecast takes your historical data and projects it forward, showing you a likely trajectory based on past performance.

It's a great tool for getting a quick gut check on where things are headed. For example, are your monthly recurring revenue (MRR) trends pointing toward hitting your quarterly goal, or are they falling short?

How to Add a Forecast to Your Report

Once you've created a basic time-series visualization (like a line chart showing website sessions over the last 90 days), adding a forecast is just a few clicks:

1. Create your visualization in a Looker Explore. Make sure you have a date/time dimension on one axis and a metric on the other.
2. Once the chart is rendered, click the gear icon to open the Edit menu.
3. Go to the Forecast tab.
4. Check the box to enable forecasting.
5. You can adjust the Length (how many periods into the future to predict) and other settings like Prediction Interval (to see a best/worst-case range).
6. Click Run, and Looker will add the projected trend to your chart, often as a dotted line.

This built-in feature is perfect for simple, linear projections. However, for more complex questions or to understand why a trend is happening, you'll want to level up to Looker's more advanced generative AI capabilities.

Going Deeper with Generative AI in Looker

With the integration of Google's AI models, Looker is becoming much more conversational. Instead of clicking through menus and manually building reports, you can now use natural language to ask questions, generate charts, and get automated explanations for what you're seeing in the data.

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

Create Reports by Asking Questions

This is where things get really powerful. In newer versions of Looker, you can describe the report you want in plain English, and the AI will build it for you. This functionality dramatically lowers the learning curve and speeds up the entire analysis process.

You can ask questions like:

• "Create a monthly trend report of our total revenue from Shopify this year."
• "Show me a line chart of new user signups from Google Ads vs. Facebook Ads for the last 60 days."
• "What was our weekly churn rate in Q2?"

Looker's AI parses your request, identifies the relevant dimensions and measures from your LookML model, and produces the visualization. From there, you can continue to refine it using chat or the standard visualization editor.

Dig for Insights with Follow-Up Questions

The real magic happens when you treat the AI like an analyst you can talk to. Once you have a trend chart, you can ask follow-up questions to understand the 'why' behind the numbers.

For example, if you see a big spike in website traffic on your chart, you can ask:

"What caused the traffic spike in the third week of July?"

The AI will analyze related data and might come back with an insight like: "The traffic increase correlates with the launch of the 'Summer Sale' email campaign, which drove an additional 5,000 sessions from email referrals." This saves you from having to manually cross-reference campaign data yourself. It does the data-digging for you.

Automating Complex Calculations

Ever get stuck trying to create a specific custom metric, like a 'week-over-week growth rate' or a lead-to-conversion rate? Generative AI can help write the LookML or formulas you need.

You can simply describe the calculation you want, such as:

"Help me create a calculated field for cost per acquisition."

The AI will generate the necessary code based on your existing ad_spend and conversions metrics, which you can then add directly to your model. This is incredibly helpful for both non-technical users who don't know the syntax and experienced analysts who want to save time.

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

Example: Building a Sales Trend Report from HubSpot Data

Let's walk through building a practical report. Imagine you're a sales manager and want to understand your team's deal creation trends and predict performance for the next month.

1. Connect Your Data: First, ensure your HubSpot data is properly modeled in Looker. Your data model should include dimensions like Deal Create Date and Sales Rep, and measures like Deal Amount and Number of Deals.
2. Ask Your Initial Question: Use the natural language feature and ask: "Show me a weekly trend of new deals created this quarter, broken down by sales rep." Looker's AI will generate a line or bar chart showing each rep's deal creation trend over the last three months.
3. Spot an Anomaly: You notice that one rep, Jane, had a massive spike in created deals three weeks ago, while another rep, John, showed a significant dip.
4. Ask Follow-Up 'Why' Questions:

• For Jane: "Why did Jane's deal creation spike in the second week of August?" The AI might find a correlation: "This spike correlates with a high volume of MQLs from the 'Annual Tech Webinar' campaign that were assigned to Jane."
• For John: "What caused John's dip in performance last month?" The discovery might be simpler: "John was on paid time off for 10 business days in August."

1. Forecast the Future: Finally, you want a forward-looking view. You can refine your prompt: "Create a single line chart showing the total number of deals created company-wide this quarter and add a forecast for the next 4 weeks." This gives you a clear vision of whether you're on track to hit your team's quarterly target.

In minutes, you've gone from raw data to a comprehensive trend report that explains past performance and projects future results - all by having a conversation with your data.

Best Practices for Better AI Trend Reports

• Be Specific: The more specific your initial prompt, the better the result. Instead of "show sales," try "show me monthly sales revenue from the US and Canada for the last 12 months."
• Iterate Your Questions: Don't expect the first chart to be perfect. Use the AI's output as a starting point and refine it with follow-up questions. "Change this to a weekly view," or "show only products in the 'Apparel' category."
• Combine AI with Human Insight: AI is a powerful tool for spotting correlations, but you provide the business context. The AI might tell you sales went up after a campaign, but you know that a competitor just went out of business, which might be the real reason.
• Sanity-Check the Data: Always confirm the AI is pulling from the correct data source and using the right dates. If an insight seems too good to be true, double-check the underlying query before sharing it.

Final Thoughts

Creating trend reports in Looker has evolved from a manual, click-intensive process to an interactive conversation. By leveraging built-in forecasting and new generative AI features, you can identify patterns, diagnose issues, and predict future outcomes faster and more easily than ever before.

These kinds of AI-powered features are streamlining how reporting gets done, shifting the workload from tedious manual data-pulling to more strategic analysis. At Graphed, we've built our entire platform around this idea. We use a conversational AI approach to help you connect all your marketing and sales data sources - like Google Analytics, HubSpot, Salesforce, and a dozen others - and create real-time dashboards and reports simply by describing what you need in plain English. This eliminates the steep setup and learning curve so anyone on your team can get immediate answers from their data.