Audience File Requirements/Schema
Accepted File Types
Accepted file types / file extensions | tsv, csv, psv, txt All file types must have the appropriate delimiter |
|---|---|
Accepted file delimiters | pipe, comma, tab |
File Format Requirements
Must include at least one primary attribute (i.e. personal unique identifier) per row (e.g. Email Address, Phone Number, etc.)
Alli currently only accepts Email or Phone as a primary identifier but may be flexible for unique cases
Please note: If your list includes more than one key attribute, the first key attribute listed will be used as the primary one.
A consistent and standardized date format must be used throughout the file
We recommend ISO format
All file should be as complete as possible with every value being filled in where available to ensure the creation of complete audience segments
File encoding is limited to UTF-8, UTF-16, or ASCII
You may drop GZ files for larger data drops
At least 1000 unique customers per file is recommended as we typically see a 50-70% match rate within ad partner platforms
Data may be hashed in SHA-256 or sent in plain text
More info on Alli security practices here Audience One Sheeter
File may contain a header to clearly outline what data a file contains OR if no header is available, we will need to know what data the file contains in order to process it correctly
Increasing the amount of PII typically increases match rates in third-party platforms as we are able to give more information about a person this way and more closely mimics the original audience we process.
File Header Requirements
The column headers for your audience list should exactly match the name and order of headers in your file to avoid import errors.
Email is the most commonly use primary attribute for audience matching in publisher platforms and should be listed first in your file and audience list. Any internal ID columns should be dropped from your file or not imported to Audiences since they cannot be used in publisher platforms and sometimes create import errors.
Column headers with 2+ words should use underscores (e.g. Purchase_Date) or combine the words (e.g. PurchaseDate) to avoid import errors.
Acceptable Audience Types
CRM lists
This is your customer summary list(s). One person per row with the row being data on that person/their purchasing habits
Transaction lists
This contains purchase information. One transaction per row that eventually gets aggregated per person
Data lists
Any non-crm and non-transaction file that can be used to map from one ID value to another within separate files (think of a join on a CRM or transaction list with additional information)
Sample File
Here is an example of how to format your CRM file.