Transformation actions help with field mapping while setting up a connection. Note that not all transformations are available for all data sources.
How transformation works
Some transformations are automatically recommended. For instance, you may notice recommended compute hash transformations on automatically mapped hashed_email and hashed_phone_number fields. You can modify the automatically recommended transformation or add additional ones as needed. You can specify a series of transformation actions on an input field, and they’ll be applied from top to bottom.
Before you begin
To apply transformations during field mapping, click Actions > Transform to view a list of available transformations that you can apply to the input field value in the pipeline before it’s imported to your Google Ads account.
Available transformation actions
Smart hash
To keep your data secure, private customer data that you import should be hashed. Data Manager will hash the data for you using the SHA256 algorithm, which is the industry standard for one-way hashing. The result is hex encoded. You don’t need to pre-format your data—Data Manager will normalize relevant PII fields, perform hashing and encoding for you, and push the data to the API for your use cases.
If you prefer to hash private customer data yourself, see Format your customer data file to ensure it’s formatted correctly. If you upload a hashed data file, don’t hash non-private customer data. Data Manager will push your hashed data to the API.
Note that smart hashing is automatic, meaning you will not need to select anything from the Actions menu.
Multiply
Multiply the column by a constant floating number.
Eligible source field types: Any numerical type.
Transform case
Change the column to uppercase, lowercase or titlecase. The transformation action will convert it to a string first.
Eligible source field types: Any type.
Merge fields
Combine two columns with a separator. This transformation will append the second value to the source field value after a specified separator string.
The format is {source_field}{separator}{column_to_merge}.
Eligible source field types: Any type.
Split
Split a column per delimiter and extract one part.
- Delimiter must be a single character. A space is a valid delimiter.
- Occurrence indicates the kth occurrence of the delimiter, where the string is to be split, beginning from “1”.
- Before is a boolean indicating whether to extract the part before the delimiter (true) or after the delimiter (false).
Eligible source field types: String.
Example:
If the input column has a value of "FirstName,MiddleName,LastName":
| Delimiter | Occurrence | Before | Result |
|---|---|---|---|
|
, |
1 | true | “FirstName” |
| , | 1 | false | “MiddleName,LastName” |
| , | 2 | true | “FirstName,MiddleName” |
| , | 2 | false | “LastName” |
Cast type
Cast the value of a column to another compatible type.
| Target type | Eligible source types | Notes |
|---|---|---|
| String | Any | Any type can be cast to string. |
| Real number |
String Integer |
Real number is of data type double If a string value can't be parsed as a double at import time, it results in an error in the "Runs" table. |
| Integer |
String Real number Percent |
If a string value can't be parsed as a double at import time, it results in an error in the "Runs" table. |
Parse timestamp
Convert a string representation of a timestamp to an integer. Date and time objects must be represented as strings in your data source and then transformed using the Parse timestamp transformation action. To learn about supported formats for time and date strings, see Prepare your data for import.
Eligible source field types: String.