type
Names a type, such as number, text or date, so you can tell Power Query what a value or column holds.
type typeExpressionReturns: A type value, which Power Query uses in type steps and in function arguments.
What it does
The type keyword writes a type. You use it in a step that sets column types, such as Table.TransformColumnTypes, and as the third argument to Table.AddColumn, so that the new column is typed when it is created. Types also appear after as in a function's arguments. Setting types is a basic habit: it makes the query fail early when data is wrong, instead of later.
Syntax and arguments
type number,type text,type date,type datetime,type logical: the basic types.type nullable number: a number that can also be null. Columns with blanks usually need this.type table [Name = text, Amount = number]: a table type with named columns.
Example
A bank export has text for the date and amount. The query sets the types once, near the end, so that the amounts add up and the dates sort.
let
Source = Table.FromRecords({
[Date = "2026-03-31", Amount = "1250", Memo = "Rent"],
[Date = "2026-03-31", Amount = "480", Memo = "Supplies"]
}),
Typed = Table.TransformColumnTypes(Source, {
{"Date", type date},
{"Amount", type number}
}),
WithFlag = Table.AddColumn(Typed, "Large", each [Amount] >= 1000, type logical)
in
WithFlag
After the type step, Amount is a number, so List.Sum and sorting work, and Date is a real date. The new Large column is typed as logical when it is created.
Example: a second, different use
Declaring a function argument as optional and nullable: (rate as nullable number) as number => rate ?? 1 accepts a null rate and uses 1 in its place.
Things to watch
- A type step fails when a value does not convert. Clean the text first, or use try to set bad values aside.
- Set the types near the end of the query, after the cleanup steps.
- Use
nullablefor columns that can be blank. Without it, a null can cause an error. - Type conversion follows the locale of the PC. For dates and decimals written in another style, pass a culture as the third argument of Table.TransformColumnTypes.
In business terms
Typed columns are the difference between a total that works and one that quietly adds text. A type step is also a small piece of documentation: it states that the amount is a number and the date is a date.