Filters in Tableau work in layers, starting from Extract Filters (which limit data before Tableau even loads it) all the way down to Table Calculation Filters (which only affect the display).
Here is the standard Tableau view of the order that filters operate in:
Here’s the execution order:
1️⃣ Extract Filters – Cut down your dataset before it even reaches Tableau.
2️⃣ Data Source Filters – Apply constraints at the connection level.
3️⃣ Context Filters – Control when other filters run.
4️⃣ Dimension & Measure Filters – Standard filters that remove data from the view.
5️⃣ Table Calculation Filters – Hide marks without filtering raw data.
The biggest mistake? Assuming that all filters work the same way—they don’t!
Let’s be clear – when we say “Table Calc Filters” what we really mean is an ad-hoc or calculated field that contains one or more Table Calculations within it.
(And THAT, is when Table Calc’s get REALLY interesting…)
Extract & Data Source Filters: The First Gatekeepers
🚀 When performance matters, start here.
Before Tableau even sees your data, you can reduce its size using Extract Filters (if using an extract) or Data Source Filters (for both live and extract connections).
✔ Use Extract Filters when working with large datasets to reduce load times.
✔ Use Data Source Filters when you want to restrict access to sensitive data—such as limiting financial records to specific departments—without impacting dashboard performance or user experience.
💡 Example Use Case: If you only need UK sales data, applying an Extract Filter means Tableau won’t even load non-UK records.
📌 Suggested Diagram: A side-by-side comparison of Extract vs. Data Source Filters, referencing [0313 Order of Operations 3.txt]0313 Order of Operation….
Key Takeaways:
- Use Extract Filters to reduce data volume before loading into Tableau—great for performance optimisation.
- Use Data Source Filters when you need row-level security or dynamic filtering within Tableau without altering the dataset permanently.
Context Filters: The Key to Unlocking Filter Control
One of Tableau’s biggest headaches is when filters don’t seem to apply in the right order.
🚩 Scenario: You want to show Top 10 Products within a Segment, but Tableau is ranking all Products across all Segments and then filtering on the chosen Segment instead.
💡 Fix: Use a Context Filter for Segment first, so Top N is applied after filtering the right subset of data:
How to Create a Context Filter
1️⃣ Right-click a filter and select “Add to Context”
2️⃣ The filter turns grey, meaning it will now run before other filters
3️⃣ All other filters now apply after the context is set
🚀 Pro Tip: Context Filters optimise performance by reducing query load – but use them wisely!
Dimension & Measure Filters: The Standard Workhorses
Once your data is in Tableau and Context Filters are applied, Dimension & Measure Filters kick in.
✔ Dimension Filters – Remove categorical data (e.g., only show “Electronics”).
✔ Measure Filters – Remove numerical values (e.g., only show Sales > £10,000).
🚩 Common Mistake: Not realising that LOD Calculations ignore Dimension Filters unless they’re in Context!
💡 Fix: Convert Dimension Filters to Context Filters if your FIXED LOD isn’t responding as expected.
Whether you need to put your dimensional filter into context to impact the Fixed Level Of Detail expression clearly depends very much on your use-case.
In the below example we might want to benchmark against all regions and therefore not apply the filter to our benchmark calculation. In this case we don’t want to be adding context. As you can see on the right hand side, the benchmark equals the same values when the filter is added to context.
The calculation in this case is:
Table Calculation Filters: The Illusion of Filtering
Table Calculation Filters don’t actually remove data—they effectively hide it from the view.
As mentioned previously, when we say a “Table Calculation Filter” what we mean is either an ad-hoc calculation or a calculated field that includes one or more table calculations and that is being used as a filter.
📌 Use Case: You apply a Year over Year Growth calculation and want to filter out empty months. A Dimension Filter would break the calculation, but a Table Calculation Filter (hiding nulls) keeps it intact.
✔ Use Table Calculation Filters when you need to control visibility without affecting calculations.
✔ Avoid them if you need actual data exclusion—they don’t reduce data size.
In the below on the left-hand side we see an empty column for the year 2021. On the right hand side we remove this by adding the Table Calculation for the growth values into the filter. See the captions for more info.
Here is the calculation for the growth.
This is then also added to the filter with the following setting:
Final Thoughts: When in Doubt, Follow the Order!
The key to debugging any issues or incorrect values you have is to think through the query flow and the order that filters are being applied:
Extract & Data Source Filters – these keep things fast
Context Filters help control the filter sequence
Dimension & Measure Filters for standard filtering
Table Calculation Filters to hide data, without removing it from the calculations
Up Next: LOD Calculations vs. Filters – Why FIXED Doesn’t Always Mean Fixed!
🚀 Ready to Level Up? Check out our next training course for Tableau Training on Tap Level Up where we deep dive into this topic.
Don’t forget to check out the previous post in this series Why Order of Operations Matters!
