How to Create a Visual Report in Tableau with AI

Building a visual report in Tableau can feel like learning a whole new language, filled with shelves, pills, and complex calculations. But Tableau is infusing its platform with AI features designed to lower that barrier, helping you create insightful reports by simply asking questions. This article will show you exactly how to use features like Ask Data and Einstein Copilot to build visual reports in Tableau, moving you from raw data to a finished dashboard much more quickly.

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 Are Tableau's Built-in AI Features?

Tableau's ecosystem includes several layers of AI meant to simplify different parts of the analytics process. Think of them as specialized assistants, each with a specific job. While different features are available depending on your license and setup (Tableau Desktop, Cloud, etc.), the core concepts are consistent.

Ask Data

Ask Data is essentially a natural language search bar for your data source. A user can type a plain-language question, and Tableau will attempt to answer that question by generating a data visualization. It's a fantastic starting point for exploring your data without needing to drag and drop fields onto a worksheet manually.

• Best for: Quick, exploratory analysis and creating single visualizations from a single data source.
• Example: Typing "total sales by country last year" and getting a map colored by sales volume.

Explain Data

Explain Data helps you understand the why behind a specific data point. When you see a spike, a dip, or an outlier in your visualization, you can use Explain Data to have Tableau automatically analyze the underlying data and surface potential explanations. It runs statistical models behind the scenes to find the most likely drivers of that value.

• Best for: Investigating anomalies and uncovering root causes without manually building dozens of different views.
• Example: Clicking on a month with surprisingly low revenue and having Explain Data tell you it was because of poor performance in a specific product category.

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

Einstein Copilot (for Tableau)

As part of the broader Salesforce ecosystem, Einstein Copilot is Tableau's generative AI assistant. It goes beyond simple queries and acts as a conversational partner in your analysis. You can ask it to perform multi-step tasks, suggest calculations, or even build entire dashboards from a description. This is the most recent and powerful addition to the AI toolkit.

• Best for: Guided analysis, complex calculations, and accelerating the dashboard-building process from start to finish.
• Example: "Create a customer analysis dashboard showing sales over time, a breakdown of sales by customer segment, and a scatter plot of profit vs. sales."

Why Bother Using AI for Tableau Reporting?

You might be a Tableau professional who is perfectly comfortable with the drag-and-drop interface, so why introduce AI? For many organizations, the traditional reporting process is a bottleneck. It often looks something like this: a marketer downloads a few CSVs on Monday, spends hours cleaning them up in a spreadsheet, builds a report for a Tuesday meeting, gets a dozen follow-up questions, and spends Wednesday finding those answers. By then, half the week is gone.

Tableau’s AI aims to break that cycle by:

• Accelerating Time to Insight: Instead of spending 30 minutes clicking, filtering, and joining to build a chart, you can get a powerful starting point in 30 seconds with a simple query.
• Empowering Non-Technical Users: Not everyone is a data analyst, but everyone needs to make data-driven decisions. AI features make data accessible to less technical team members, allowing them to ask questions and find answers without relying on a dedicated data person.
• Uncovering Hidden Patterns: It's easy to miss subtle correlations by manually slicing and dicing data. Features like Explain Data automatically scan for relationships, helping you find insights you weren't even looking for.
• Reducing Manual Workload: AI handles the tedious parts of report creation, freeing you up to focus on the strategic part: interpreting the results and deciding what actions to take.

Step-by-Step: Creating a Report with Tableau's AI Features

Let's walk through building a simple sales performance report using Tableau's AI tools. For this example, we’ll imagine we're working with a standard e-commerce dataset that includes fields like Order Date, Sales, Profit, Category, State, and Product Name.

1. Prepare Your Data and Access "Ask Data"

Before you can use any AI tools, Tableau needs to know what it's working with. The effectiveness of any AI conversation is directly tied to the quality of the data source it’s using.

• Connect to Your Data: Start a new Tableau workbook and connect to your data source (e.g., an Excel file, a database connection, a Google Sheet).
• Curate the Data Source (Important!): For "Ask Data" to work well, Tableau needs to "index" the data source. When publishing your data source to Tableau Server or Tableau Cloud, ensure you select the "Ask Data" option. Sometimes, you might need to build an "Ask Data Lens," which is a curated version of your data source tailored for specific audiences. You can rename fields to be more intuitive (e.g., changing "Cust_Seg_ID" to "Customer Segment") and add synonyms. This upfront curation helps the AI understand your prompts better.

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

2. Start Your Analysis with "Ask Data"

Your goal is to build a high-level overview of sales. Let's start by looking at sales performance over time.

In the "Ask Data" interface for your published data source, simply type in a question:

Show month-over-month sales for this year

Ask Data will interpret this request. It understands "sales" is your key metric, "month-over-month" implies a line chart over a date dimension, and "this year" is a filter. It will automatically generate a line chart showing your sales trends. You can save this visualization as a new worksheet to include in your dashboard later.

Now, let's create a geographic view:

Sales by state as a map

Tableau recognizes "state" as a geographic field and "map" as the desired visualization type, creating a filled map of the U.S. where darker colors indicate higher sales. Save this worksheet as well.

3. Dig Deeper with "Explain Data"

Looking at your sales map, you notice that California has significantly higher sales than any other state. That's an outlier worth investigating. Why is it performing so well?

• Click on the California mark (the shaded state on the map).
• In the tooltip that appears, you’ll see a lightbulb icon. This is the entry point for "Explain Data." Click it.
• A new window opens, and Tableau’s AI gets to work. It runs models to identify potential drivers and presents them as a series of supplemental visualizations.

You might see an explanation that suggests California’s high sales are driven by strong performance in the "Technology" and "Office Supplies" categories. Another explanation might show you a histogram of per-order sales, revealing more high-value orders in that state. This is an insight that would have taken several manual steps to uncover. You can select any of these generated explanations and open them as new worksheets to add to your report.

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

4. Assemble the Dashboard with Einstein Copilot

Now you have a few individual charts, but you still need a dashboard. This is where a more advanced assistant like Einstein Copilot comes into play. If your version of Tableau includes this feature, you'll see an Einstein icon to open a conversational panel.

Here you can make more complex requests that involve multiple worksheets or actions:

Create a new dashboard named "Executive Sales Overview". Add the "Sales by State" worksheet and the "Monthly Sales Trend" worksheet. Also, add a bar chart showing the top 10 products by profit.

Einstein Copilot will process this multi-step request:

• It creates a new dashboard.
• It pulls in the two worksheets you already created.
• It generates a new worksheet with the top 10 products by profit and adds it to the layout.

You now have a first draft of your dashboard built from a single command.

5. Refine and Iterate

Your dashboard is functional but might need some polishing. You can continue using a mix of AI commands and manual clicks to refine it.

• Add Interactivity: You can ask Einstein Copilot, "Make the map a filter for the other charts." This will make it so that clicking on a state in the map will filter the monthly trendline and top product chart to show data only for that state.
• Adjust Formatting: You can then manually adjust colors, font sizes, and the layout to match your company's branding, or use prompts like "Change the color palette to the 'Summer' theme."
• Add Titles and Text: Finally, add a clear title and text boxes with commentary to provide context for anyone reading the report.

Final Thoughts

By leveraging tools like Ask Data and Einstein Copilot, you can drastically cut down the time it takes to build a purposeful visual report in Tableau. Using plain-language prompts for initial exploration and relying on automated analysis for deeper insights lowers the technical barrier, allowing anyone on your team to move from raw data to actionable information more efficiently.

While the AI in powerhouse tools like Tableau is becoming incredibly sophisticated, it's still part of a complex platform that often requires data engineers and significant setup. For marketing and sales teams who just need answers fast without a massive BI infrastructure, we built Graphed. We focused on making AI do all the work - from connecting to your specific tools like Google Analytics, Shopify, and Salesforce in one click, to instantly building live dashboards from a single sentence. Instead of navigating layers of menus, prompts, and server settings, you just ask a question like "what's our campaign ROI?" and get a shareable, real-time dashboard seconds later.