What is Tableau MCP?

Curious about what a Tableau Multiconnection Data Source or an MCP file is? You’ve come to the right place. While the term “MCP” can refer to a few different things in the Tableau ecosystem, what most analysts are searching for is a way to stop jumping between isolated data sources and start analyzing them together. This article breaks down exactly how to connect, federate, and save your diverse data sources - like Salesforce, Google Analytics, and various spreadsheets - into a single, reusable resource so you can start getting the complete picture of your business performance.

Graphed

Your AI Data Analyst to Create Live Dashboards

Connect your data sources and let AI build beautiful, real-time dashboards for you in seconds.

Build Your First Dashboard

Demystifying Tableau's Multi-Connection Data Sources

First, let's clear up the terminology. While you might occasionally run into niche file types, when people talk about combining multiple data connections in Tableau, they're almost always referring to creating and saving a Tableau Data Source (.tds) file or a Tableau Packaged Data Source (.tdsx) file. These files are the heart of efficient, collaborative analytics in Tableau.

Think of it like this:

• A .tds file is like a recipe. It doesn't contain the actual ingredients (your raw data). Instead, it holds all the instructions: where to find the data sources (e.g., your SQL server login, your Google Sheets link), how the tables should be joined or related, any custom calculations you've created, and formatting you've applied. It's a small, portable file that acts as a shortcut to your perfectly configured data model.
• A .tdsx file is a meal kit. It contains both the recipe (.tds) and all the prepped ingredients (a snapshot, or .hyper extract, of your data). This packaged file is completely self-contained, makes dashboards load incredibly fast, and is perfect for sharing with colleagues who don't have access to the live databases.

Why Should You Bother with a Multi-Connection Setup?

Wrestling data from different platforms is a massive time sink. Your ad data is in a Google Sheet, your sales data lives in Salesforce, and your website traffic is in Google Analytics. Creating a multi-connection data source centralizes this effort, turning hours of repetitive work into a one-time setup with four key benefits.

AI for Data Analysis

Crash Course

Graphed

Free PDF Guide

AI for Data Analysis Crash Course

Learn how to get AI to do data analysis for you — the best tools, prompts, and workflows to go from raw data to insights without writing a single line of code.

Download Free Guide

1. Create a Single Source of Truth

When everyone on your team connects to data separately, you get chaos. One person might join tables differently, another might name a calculated field "Return on Ad Spend" while someone else calls it "ROAS." By creating a single, shareable .tds file, you establish the official data model. All of the business logic - joins, relationships, key calculations like conversion rate, customer lifetime value, or channel-specific ROI - is baked right in. Every report built from this source is automatically consistent.

2. Break Down Pesky Data Silos

The most powerful insights come from combining data, not viewing it in isolation. A multi-connection data source is the bridge between your operational silos. When you connect your campaign spend from Google Ads to your sales outcomes in your eCommerce platform, you don’t have to wonder if a campaign performed well - you can see exactly how much revenue was generated for every dollar spent, all in one view. This is how you discover which marketing channels are actually driving the business forward.

3. Drastically Speed Up Your Workflow

The typical reporting cycle is a painful, manual process: download several CSVs on Monday, wrestle them into one master spreadsheet, build your pivot tables for a Tuesday meeting, and then spend Wednesday answering follow-up questions. A saved data source blows up this process. Connect your sources, model them once, and save your work as a .tds file. From then on, all you have to do is connect to that one file to access all your blended, up-to-date data. Better yet, when the data is published to Tableau server, workbooks can stay on an automated refresh schedule.

Graphed

Your AI Data Analyst to Create Live Dashboards

Connect your data sources and let AI build beautiful, real-time dashboards for you in seconds.

Build Your First Dashboard

4. Empower Your Entire Team

Not everyone on a marketing or sales team is a data professional, nor should they have to be. With a saved multi-connection data source, one technical person can handle the complexities of setting up the joins and creating the necessary calculations. Then, the rest of the team - the channel managers, analysts, social team, and content creators - can connect to that pre-built data source. The team doesn't have to stress about writing custom code - they simply drag and drop from a clean, reliable, and logical data set and find insights without waiting in line for help from another team member.

Working With Multiple Data Sources: Relationships vs. Joins vs. Blending

"Connecting" to multiple sources is just the first step. You then need to tell Tableau how these different data sets talk to each other. Tableau offers three primary methods for this, each with its own specific use case.

Relationships (The Modern Go-To)

Tableau Relationships are the new default, introduced as part of Tableau's "noodle" data model. They are flexible, smart, and efficient. Instead of physically mashing two tables together into one massive, new table, you simply define the fields that relate them (like Order ID or Customer Email).

• How they work: You tell Tableau how two tables are connected, but Tableau keeps them separate. When you build a visualization, Tableau intelligently queries only the necessary data from each table at the correct level of detail, minimizing data duplication and avoiding common aggregation errors.
• Use them when: This is your best option in almost all modern scenarios. Use relationships anytime you have related data in different tables, especially if those tables are at different levels of granularity (e.g., daily ad spend and individual customer orders).

Joins (The Classic, Rigid Approach)

If you've ever worked with SQL, you already understand joins. A join physically combines two or more tables into a single, wide table by matching rows based on a common field. You effectively create a new, fixed dataset before you even start building a chart.

• How they work: You define the join type (Inner, Left, Right, Full Outer) and the join key. Tableau then merges the data, and every worksheet in your entire workbook will use this newly created combined table.
• Use them when: Use joins when you are 100% certain you only and always need the data combined that way for the scope of the project and want the tables row-wise ahead of time. It's less flexible than relationships but can be useful for simpler models - just be wary of accidentally duplicating data if your tables are at different levels of detail.

Data Blending (The Final Option)

Data blending is fundamentally different from joins or relationships. It's a view-level integration, meaning Tableau queries each data source independently and then aggregates the results. It then "blends" - or links - those results on a common field within your visualization.

• How they work: One data source is designated as the "primary," and any others are "secondary." The secondary data source can only provide data at the level of aggregation defined by the primary linking field in your view. You'll recognize a blend by the little orange broken-link icon next to a data source.
• Use them when: Use blending sparingly, typically when you cannot use a relationship or join. The most common use case is when your data lives in different places that can't be federated in a single federated connection (e.g., combining live Salesforce data with a local Excel file) or when working with published data sources you can't edit.

AI for Data Analysis

Crash Course

Graphed

Free PDF Guide

AI for Data Analysis Crash Course

Learn how to get AI to do data analysis for you — the best tools, prompts, and workflows to go from raw data to insights without writing a single line of code.

Download Free Guide

Step-by-Step: Creating Your Reusable Data Source File

Ready to build your first .tds file so you never have to connect your Google Analytics and Shopify data from scratch again? It’s easier than you might think.

1. Connect to Your First Source: Open Tableau Desktop and connect to your primary data source, like a PostgreSQL database or a Google Sheet.
2. Connect to Additional Sources: In the Data Source pane, click the "Add" button next to Connections. Now, connect to your second data source, like Google Analytics or Salesforce Accounts. You will see both data sources listed in the left pane.
3. Establish the Connections: Drag-and-drop your tables from both connections into the main canvas. Tableau will often automatically create a "noodle" (a relationship) between them if it detects a common field name. If not, click on the noodle to manually define which columns relate to each other.
4. Clean and Enhance Your Data Model: This is where the magic happens.
5. Save Your Masterpiece: Once your data model is perfect, navigate to the Data menu at the top. Hover over the name of your data source, and in the fly-out menu, select "Add to Saved Data Sources…"
6. Choose Your Format: A dialog box will appear. Give your data source a descriptive name (e.g., "Mkg_Sales_Master"). Now, choose your format: save it as a Tableau Data Source File (.tds) to save just the connection metadata, or a Tableau Packaged Data Source File (.tdsx) to include an extract of the current snapshot of your connected data set.

That's it! The next time you - or a teammate - wants to build a report with this data, they can simply choose "Connect to a Saved Data Source" and select your file, bypassing all the tedious setup and getting straight to finding actionable, reliable insights every time.

Final Thoughts

Learning to correctly connect, relate, and save multi-connection data sources is a major leverage point for anyone using Tableau. It’s what separates analysts stuck in the endless, weekly data prep quicksand from managers who can build consistent, powerful, and truly integrated reports efficiently, and on their own timeframes. By taking the time to build one unified data model and share it as a reusable file, you make your analytics practice more accurate and scalable.

Of course, there is a serious learning curve that comes with understanding data models, choosing between relationships and joins, and becoming proficient inside a complex tool like Tableau. We built Graphed because we believe getting insights from your data shouldn't require an eighty-hour masterclass. Instead of manually configuring connections and building reports by hand, you just connect your platforms like Shopify, Salesforce, and Google Ads once, then ask for what you need in plain English - like "create a dashboard showing ROAS by ad campaign for the last 90 days." Our AI handles the connections, modeling, and visualization, so your whole team can build live dashboards and get data-driven answers in seconds, not afternoons.