
What Does FIXED Do in Tableau?
Carlos GarciaPublished What Does FIXED Do in Tableau?
You find a calculated field in someone else's workbook that reads {FIXED [Customer ID] : SUM([Sales])} and the curly braces alone tell you this is not an ordinary calculation. There is a colon in the middle of it, a dimension named before the colon, and an aggregation after it.
That is a level of detail expression, and FIXED is the keyword that makes it behave differently from every other calculation on the sheet. It is also the single most common fix for a number that looks almost right and is quietly wrong.
The short version is that FIXED lets you compute a value at a granularity you choose, instead of the granularity the worksheet happens to be drawn at. Everything else about it follows from that one idea.
This guide covers the syntax, exactly which filters a FIXED expression ignores, where it lands in Tableau's processing order, and the cases where you should reach for INCLUDE or EXCLUDE instead.
What Does FIXED Do in Tableau? The Short Answer
FIXED computes an aggregate at the dimensions you declare and nowhere else. Tableau's documentation puts it plainly: FIXED expressions compute values using the specified dimensions "without reference to the view level of detail."
That last clause is the whole feature. A normal aggregation like SUM([Sales]) returns a different number in every view, because it aggregates up to whatever dimensions you have dropped on Rows, Columns, Color and the rest. A FIXED expression returns the same number regardless of what is in the view.
So {FIXED [Customer ID] : SUM([Sales])} is lifetime sales per customer. Put Region on Rows and it is still lifetime sales per customer. Put Order Date on Columns and it is still lifetime sales per customer, repeated across every column.
Tableau notes that a FIXED expression can sit at a finer level of detail than the view, a coarser one, or the same one. It is not pinned above or below the view. It is simply independent of it.
Not sure whether the numbers in your reporting are being computed at the level you think they are? Get a free SEO audit from SEO Stuff and find out what your dashboards are actually measuring.
Free audit
Where do you stand in AI search?
A manual SEO + AI search audit of your site: how ChatGPT, Gemini, Perplexity and Claude read and cite it today, and what to fix first. Report within 48 hours, no credit card.
Get my free auditOr have it all done: AI SEO services, $999 once
Why Does Tableau Need a Level of Detail Keyword At All?
The view decides the granularity, and that is usually fine
In an ordinary Tableau worksheet, the dimensions you place on the shelves define the level of detail, and every measure aggregates up to it. Drop Region on Rows with SUM([Sales]) and you get sales per region. Add Category and the same measure now returns sales per region per category.
This is convenient almost all of the time. It is also the reason a lot of calculations cannot be expressed at all without a level of detail keyword.
The average-of-ratios problem
The classic case is a ratio you want averaged. Suppose you want average revenue per customer, by region. The instinct is to write SUM([Sales]) / COUNTD([Customer ID]), and in a region-level view that happens to work.
Now suppose you want the average of each customer's own total, rather than the regional total divided by a customer count. Those are different numbers, and the second one cannot be built from a view-level aggregation, because the view is at region granularity and the number you need lives at customer granularity.
FIXED solves it by computing the customer-level total first, independently of the view, and letting you aggregate that result afterward. You write AVG({FIXED [Customer ID] : SUM([Sales])}) and the inner expression is evaluated per customer no matter how the sheet is arranged.
The same shape shows up constantly. Average order value when the view is by month. Percent of a total that should not move when a user filters. First purchase date per customer displayed on a row that is not per customer. All of them are a mismatch between the granularity of the question and the granularity of the view.
The three keywords, in one line
Tableau's own framing is the clearest summary available. Level of detail expressions let a calculation run at a more granular level with INCLUDE, a less granular level with EXCLUDE, or an entirely independent level with FIXED.
INCLUDE adds the dimensions you name to whatever is already in the view. EXCLUDE subtracts dimensions from the view level of detail. FIXED ignores the view entirely and uses only what you declared.
How Do You Write a FIXED Expression, Step by Step?
The structure Tableau documents is a keyword, a dimension declaration, a colon, and an aggregate expression, all inside curly braces.
- Create a calculated field. Right-click in the Data pane and choose
Create Calculated Field, or use the Analysis menu. - Open with a curly brace and the keyword. Type
{FIXED. The braces are part of the syntax, not formatting. - Declare your dimensions. Name the dimensions that should define the granularity, separated by commas:
{FIXED [Region], [Category]. Tableau allows any expression that evaluates as a dimension here, including date expressions. - Add a colon. The colon separates the dimension declaration from the calculation.
- Write an aggregate expression. This part must be aggregated.
{FIXED [Region] : SUM([Sales])}is valid;{FIXED [Region] : [Sales]}is not. Tableau also notes that theATTRaggregation is not supported and that table calculations are not permitted in the aggregate expression. - Close the brace and save. The finished field behaves like any other field in the Data pane.
A detail worth knowing: the dimension declaration can be empty. Tableau documents {FIXED : MIN([Order Date])} as equivalent to the table-scoped form {MIN([Order Date])}, which computes across the entire table. The documentation never calls the declaration "optional" in those words, but it shows the empty form as valid.
Whether the result is a measure or a dimension depends on the field you aggregated. Tableau states that a FIXED expression can produce either, while INCLUDE and EXCLUDE expressions are always measures.
Reporting that looks correct and measures the wrong population is the most expensive kind of broken. Request a free audit and get a second pair of eyes on yours.
Which Filters Does a FIXED Expression Ignore?
This is where FIXED surprises people, and it is worth memorizing rather than deriving.
Tableau's documentation gives the rule as a short exclusion list: FIXED expressions ignore all the filters in the view other than context filters, data source filters, and extract filters.
Read that backwards and it is more useful. Three kinds of filter still reach a FIXED expression. Everything else on the Filters shelf does not.
So an ordinary dimension filter on Region will change your SUM([Sales]) and leave {FIXED [Customer ID] : SUM([Sales])} completely untouched. Tableau's own worked example makes the same point with a percent-of-total calculation: a state filter changes the numerator and not the FIXED denominator.
That behavior is sometimes exactly what you want. A percent-of-total that holds its denominator while the user filters is a genuinely useful thing, and it is hard to build any other way.
When it is not what you want, Tableau documents two fixes. Promote the relevant filter to a context filter, using Add to Context on the Filters shelf, which puts it in the small set of filters a FIXED expression respects. Or rewrite the calculation as an INCLUDE or EXCLUDE expression, which Tableau explicitly suggests as the alternative to using context filters.
Where Does FIXED Sit in Tableau's Order of Operations?
Tableau publishes an order of operations page, and this is the honest state of it: the full ordered stage list appears there only as a flowchart image. The page introduces the list and then shows a diagram. There is no prose version of the complete sequence on either that page or the level of detail overview.
Plenty of blog posts reproduce a full ordered list anyway. This one will not, because the only text the vendor actually commits to is a handful of pairwise statements. Those are worth having exactly:
- FIXED calculations are applied before dimension filters. The level of detail overview and the order of operations page both state this.
INCLUDEandEXCLUDEexpressions are considered after dimension filters. That is the structural difference between them andFIXED.- Context filters apply per worksheet and run before any other filter you create in a worksheet.
- A latest-date filter is set after data source filters and before context filters.
- A dimension filter is applied before a table calculation.
- Table calculation filters are applied after calculations run, so they hide marks without filtering the underlying data used in those calculations.
- General and top-N dimension filters sit at the same place in the order.
What the text does not state is where FIXED lands relative to extract filters, data source filters, measure filters or table calculations. The exclusion list in the previous section implies extract and data source filters come first, but the order of operations page does not spell it out in words.
The one relationship you need for daily work is the first bullet. FIXED runs before dimension filters, which is the mechanical reason it ignores them.
Getting the order of operations right in a dashboard is the same discipline as getting it right in your analytics. Claim a free SEO audit and see which of your numbers are computed on the wrong population.
When Should You Use FIXED?
Reach for FIXED when the granularity of your question is fixed by the business logic rather than by the layout of the sheet.
Per-entity totals shown in a different view. Lifetime value per customer, total hours per project, revenue per vehicle. The entity is the grain, and the view is going to be something else.
A denominator that must not move. Percent of total against a stable base, cohort size, share of company-wide revenue. This is the case where ignoring dimension filters is the feature rather than the bug.
Ratios of totals rather than averages of ratios. Compute the numerator and denominator at the grain where they make sense, then aggregate the ratio.
First or last event per entity. {FIXED [Customer ID] : MIN([Order Date])} gives an acquisition date you can then bucket, compare or subtract from, in a view that is not per customer.
Cross-grain comparisons. Anything of the form "this row's value against a total computed somewhere else" tends to be a FIXED expression plus arithmetic.
If your requirement instead reads as "the same calculation, but also broken out by X" then INCLUDE is the better fit. If it reads as "the same calculation, but ignoring X" then EXCLUDE is.
What Are the Limitations of FIXED?
Some data sources do not support level of detail expressions at all. Tableau's support table lists Actian Vectorwise, multidimensional cubes, DataStax Enterprise, Microsoft Access and Microsoft Jet-based connections, and Splunk as not supported. Generic ODBC is listed as limited and dependent on the specific source, and Google BigQuery is supported for standard SQL but not legacy SQL. Several other databases carry minimum version floors.
Complexity can bite. Tableau's note is that it will not disable these calculations for the affected databases, but that query errors are possible if calculations become too complex. That is the only performance-adjacent warning the documentation gives.
A few smaller constraints are documented and easy to trip over:
- Floating-point measures can behave unreliably when used in a view that requires comparison.
- Level of detail expressions are not shown on the Data Source page.
- A
USERNAME()function is removed when a workbook is created using that data source. - In a dimensionality declaration, reference a parameter by name rather than by value.
- With data blending, the linking field from the primary data source must be in the view.
EXCLUDEcannot be used in row-level expressions.
One more absence worth naming. None of Tableau's three relevant pages says which Tableau version introduced FIXED or level of detail expressions generally. If you need that for a compatibility decision, the documentation will not give it to you.
How Does FIXED Compare to INCLUDE and EXCLUDE?
The practical difference is what each one does with the dimensions already in your view.
FIXED disregards them. You get exactly the dimensions you declared, which makes the result stable across every view you drop it into and makes it the right choice when the grain is a business fact.
INCLUDE adds your declared dimensions to the view's own. The result is more granular than the view and then aggregates back up, which is how you average something computed at a finer grain. Tableau notes that an INCLUDE expression will never be replicated across marks.
EXCLUDE removes dimensions from the view level of detail, producing a coarser number. Tableau notes that EXCLUDE expressions always cause replicated values to appear in the view, which is normal and expected rather than a mistake.
The filter behavior is the other axis, and it is the one people forget. FIXED runs before dimension filters and ignores them. INCLUDE and EXCLUDE are considered after dimension filters and respect them. If you want a calculation that honors whatever the user filtered, that difference matters more than the granularity.
A reasonable default: start with FIXED when you can name the grain outright, and switch to INCLUDE or EXCLUDE when you catch yourself adding context filters purely to make a FIXED expression respond to the view.
Dashboards are only as good as the data arriving in them. Get a free SEO audit and find out whether your traffic numbers are telling the truth before you build on top of them.
Final Thoughts
FIXED is one keyword doing one thing: computing an aggregate at the dimensions you name, without reference to the view. The syntax is a brace, a keyword, a dimension list, a colon and an aggregation.
Almost every surprising thing about it traces back to that independence. It returns the same number in any view because it does not consult the view. It ignores your dimension filters because it is computed before they run.
The two habits that keep it out of trouble are naming the grain before you write the calculation, and checking whether you actually want the filters to apply. If you do, INCLUDE or EXCLUDE is probably the better keyword, and reaching for context filters to force the issue is a sign you picked the wrong one.
Context filters are the other half of this story, since they are one of only three filter types a FIXED expression respects. Our guide to what Add to Context does in Tableau covers that mechanism in detail, including where context filters sit in the processing order, and it pairs directly with this one.
Want this done for you?
SEO Stuff gets businesses cited by ChatGPT, Gemini, Perplexity and Claude, and ranking in Google. Start with the free audit, a call, or the package.
See where you stand
A free, manual SEO + AI search audit of your site, with a report within 48 hours. No credit card, no sales call.
Get my free auditTalk it through
A free 15-minute call with the founder about whether AI search is a meaningful opportunity for your business.
Free strategy callHave it done
The Done-For-You Package: audit, 10 pages of content, 3 DR50+ placements, dashboard. $999 once, delivered in 21 business days.
See the package

