How to Disaggregate Data in Tableau

When you create a view in Tableau, it automatically rolls up your data into tidy summaries. This is called aggregation, and it's perfect for seeing the big picture - like total sales per region or average profit per category. But what if you need to see individual trees instead of the forest? To do that, you need to disaggregate your data. This article will walk you through exactly how and why to turn off aggregation to uncover deeper patterns, outliers, and relationships in your data.

What’s the Difference? Aggregated vs. Disaggregated Data

Before jumping into the steps, it’s helpful to understand the core difference between these two views of your data. Think of it like looking at a city from a satellite versus walking its streets.

Aggregated Data (The Default View)

Aggregation combines multiple individual data rows into a single value. When you drag a measure like 'Sales' onto a Tableau sheet, it automatically becomes SUM(Sales) or AVG(Sales). Tableau is programmed to do this because most business questions involve summaries - totals, averages, counts, etc.

For example, if you drag Region to Columns and Sales to Rows, Tableau doesn't show you a bar for every single sale. It shows you four bars, one for each region, with each bar representing the sum of all sales in that region.

What it is: A summary of your data (e.g., sum, average, min, max, count)
When to use it: For high-level reporting, understanding overall performance, and creating standard charts like bar charts, pie charts, and maps.
The Result: Fewer marks on the screen, faster performance, and a clear overview.

Disaggregated Data (The Detailed View)

Disaggregation does the opposite. It displays every single row from your data source as an individual mark in your visualization. If your dataset has 10,000 rows (representing 10,000 individual sales), a disaggregated view will show all 10,000 marks.

This detailed view is essential for certain types of analysis where an average or sum can hide crucial details. Instead of seeing the total sales, you see the field of individual transactions that make up that total.

What it is: A raw view of every individual data point
When to use it: For spotting outliers, understanding data distribution, analyzing relationships between two measures (like in a scatter plot), and performing granular analysis.
The Result: Many marks on the screen, potentially slower performance, but a much richer and more detailed perspective.

Why Would You Want to Disaggregate Your Data?

Aggregation is great, but it can sometimes oversimplify your data, leading you to miss important insights. Disaggregating helps you look under the hood.

1. To See Every Data Point in a Scatter Plot

This is the most common reason to disaggregate data. Imagine you want to see the relationship between the profitability of a product and the discount it was given. An aggregated view might show you the average profit and average discount, represented by a single dot. That’s not very useful.

By disaggregating, you create a scatter plot where each dot represents a single order or product. This lets you visually identify trends, like whether higher discounts consistently lead to lower profits, or if there are clusters of highly profitable products with low discounts.

2. To Identify Outliers and Anomalies

Averages can be misleading. A region might have a healthy average profit, but this could be skewed by a few extremely profitable sales while the rest are losing money. Aggregation hides these extremes.

Disaggregating the data lays everything bare. You can immediately spot those outlier transactions - the massive sale that drove the average up, or the series of a dozen money-losing orders that are being masked by profitable ones. This helps you identify what’s working (and what isn't) at the most granular level.

3. To Understand Data Distribution

Let's say the West and East regions both have $500,000 in total sales. From an aggregated view, they look identical. But when you disaggregate, you might discover something new:

The West region achieved its total with thousands of small-value sales.
The East region achieved its total with just a few massive, high-value sales.

This context is invaluable. The sales strategy needed to grow the West region (focusing on volume) is completely different from the one needed for the East region (focusing on high-ticket clients). Disaggregation reveals this underlying distribution pattern, which would be totally invisible otherwise.

How to Disaggregate Data: A Step-by-Step Guide

Turning off aggregation in Tableau is surprisingly simple. It’s a single click, but knowing where to find it is the key.

Let's use the Sample - Superstore dataset that comes with Tableau to walk through it.

Step 1: Create a Basic, Aggregated View

First, let's see what Tableau does by default.

Open Tableau and connect to the Sample - Superstore data.
Drag the Sales measure to the Columns shelf.
Drag the Profit measure to the Rows shelf.

You’ll see just a single mark on your view. If you hover over it, you'll see it represents the total sum of sales and the total sum of profit for the entire dataset. This is Tableau’s default aggregated view - not very insightful for comparing these two measures.

Step 2: Disaggregate Your Measures

Now for the magic. All you need to do is tell Tableau to stop summing everything up.

Click on the Analysis menu in the top navigation bar.
In the dropdown menu, you'll see a checked option called Aggregate Measures. Simply click on it to uncheck it.

Instantly, the single mark explodes into a cloud of thousands of points. You're now looking at a disaggregated view - a scatter plot where each mark represents a single row from your data source (in this case, an item within an order).

That's it! You've successfully disaggregated your data.

Taking Your Disaggregated View to the Next Level

Simply disaggregating the data is a great start, but the real power comes from adding more context to your new scatter plot.

Use Color, Size, and Detail to Add Context

Now that you have individual marks, you can encode them with additional information using the Marks card.

Color: Drag a dimension like Category or Region to the Color shelf. This will color-code each point, letting you see if certain product categories or regions cluster in specific parts of the scatter plot (e.g., one category is always high-profit, while another is consistently low-profit).
Size: Drag a measure like Discount or Quantity to the Size shelf. This will make marks with higher values larger, providing another layer of visual analysis. For instance, you could quickly see if larger discounts (bigger marks) correlate with lower profit.
Detail: If you want to ensure each mark represents something specific, like a particular customer or order, drag a unique dimension like Order ID or Customer Name onto the Detail shelf. This explicitly sets the level of detail for your marks.

Use Analytics Tools like Trend Lines

Once your view is disaggregated, you can use Tableau's built-in analytics tools.

Navigate to the Analytics pane (next to the Data pane). From there, you can drag Trend Line onto your view. Tableau will instantly draw a line of best fit through your scatter plot, giving you a statistical visualization of the relationship between your two measures (like Sales and Profit).

Potential Pitfalls and Best Practices

Disaggregating data is a powerful analytical technique, but there are a couple of things to watch out for.

1. Performance Issues with Large Datasets

Remember, disaggregating means Tableau has to plot every single row of your data. If your dataset has millions or even billions of rows, trying to plot them all will seriously slow down Tableau or even cause it to crash.

Best Practice: Before you disaggregate, use filters to narrow down your data. For example, filter down to a specific year, region, or product category first. This reduces the number of marks Tableau has to render, keeping your workbook snappy and responsive.

2. Overplotting (The Big Blob Problem)

When you have thousands of marks on a view, they often overlap so much that they form a giant, unreadable blob. This is known as overplotting.

Best Practices to Fix Overplotting:

Adjust Transparency: On the marks card, click on Color and reduce the opacity. This allows you to see where marks are most densely concentrated.
Make Marks Smaller: Click on Size on the Marks card and scale down the mark size.
Filter Your Data: As with performance issues, filtering is your best friend. Zooming into a subset of your data can often reveal patterns that were hidden in the larger blob.

Final Thoughts

Moving from a high-level summary to a detailed view by disaggregating your data is a fundamental skill in Tableau. It unlocks a deeper layer of analysis, allowing you to build insightful scatter plots, find hidden outliers, and truly understand the composition and distribution of your dataset beyond simple totals and averages.

Creating these kinds of detailed views should be easy, yet sometimes it feels like a chore, clicking through menus just to ask a simple question. We built Graphed because we believe anyone should get answers from their data just by asking. Instead of manually building charts, you can ask in plain English, "Show me a scatter plot of sales versus profit for every order from our Shopify store," and instantly get a fully interactive visualization without having to worry about settings, formulas, or whether your data is aggregated.