Level of Detail Expressions

A Level of Detail (LOD) expression lets you compute an aggregation at a specific granularity — independent of the Dimensions currently in the view. Where a regular calculated field always aggregates at the view level, an LOD expression fixes the aggregation level you want regardless of what the view shows.

The Problem LOD Expressions Solve

Imagine you want to show each order's Sales alongside the average Sales for that customer — on the same row-level view. A regular SUM([Sales]) in a view grouped by Order gives you each order's total. But AVG([Sales]) also computes at order level when the view is at order level. You need the average computed at Customer level, then brought back down to Order level for comparison. A regular calculated field cannot do this. An LOD expression can.

LOD Syntax — The Three Keywords

Every LOD expression follows the same structure: a keyword in curly braces, a Dimension declaration, a colon, and an aggregation. The keyword controls how the declared Dimensions relate to the view's Dimensions.

{ FIXED [Dimension] : AGG([Measure]) }
{ INCLUDE [Dimension] : AGG([Measure]) }
{ EXCLUDE [Dimension] : AGG([Measure]) }

FIXED — Compute at an Exact Granularity

FIXED computes the aggregation at exactly the Dimensions you specify — ignoring whatever Dimensions are in the view. The result is the same regardless of what the view shows. FIXED is the most commonly used and most powerful LOD type.

Total Sales per Customer — computed at Customer level, attached to every row

{ FIXED [Customer Name] : SUM([Sales]) }

First Order Date per Customer — the customer's acquisition date

{ FIXED [Customer Name] : MIN([Order Date]) }

FIXED — Data Flow Mockup

INCLUDE — Add a Dimension to the View Level

INCLUDE computes at a finer granularity than the view by temporarily adding the specified Dimension to the view's Dimensions. The result is then aggregated back up to the view level. INCLUDE is useful when you want to compute something at row or sub-group level and then average or summarise the results in the view.

Average Order Value per Customer — average of per-customer totals shown at Category level

AVG( { INCLUDE [Customer Name] : SUM([Sales]) } )

EXCLUDE — Remove a Dimension from the View Level

EXCLUDE computes at a coarser granularity than the view by temporarily removing the specified Dimension from the calculation. This is the right choice when you need a subtotal or grand total alongside detail — such as showing each Sub-Category's Sales alongside the total for its Category, without adding a separate grand total row.

Category Total alongside Sub-Category detail — benchmark comparison

{ EXCLUDE [Sub-Category] : SUM([Sales]) }

FIXED vs INCLUDE vs EXCLUDE — Side by Side

FIXED and Dimension Filters — The Important Caveat

FIXED LOD expressions are computed before dimension filters are applied in Tableau's execution order. This means a FIXED expression will include data from rows that a dimension filter would otherwise remove. For example, a FIXED Customer Sales total will include all of a customer's orders — even if the view is filtered to show only the East Region.

To make a FIXED expression respect a filter, promote that filter to a context filter. Context filters run before FIXED LOD expressions in the execution order, so the FIXED calculation operates only on the context-filtered data. This is the most common LOD debugging step in real-world workbooks.

Next up — Lesson 39: FIXED, INCLUDE, and EXCLUDE in depth — advanced LOD patterns including cohort analysis, customer segmentation, and nested LOD expressions.