Custom Data Onboarding

This guide walks you through the steps to securely upload your customer data to Experian for audience activation to supported destinations. It provides all the details needed to prepare and format your audience files.

Who to contact:

Your Data > Secure Transfer Service (STS) > Experian Activation

Acceptable file transfer methods: 

• STS web application (browser-based upload)
• SFTP via STS
• AWS S3: Given access, STS can read audience files from an S3 bucket you own.

Ask your Project Manager for the STS User Guide for step-by-step instructions.

Refreshing an audience involves updating the individuals within it while keeping the existing audience name unchanged.

To refresh an existing audience, simply upload a new file using the exact same file name. This will completely replace the previously uploaded audience data.

Audiences expire 180 days after their creation, or after the last refresh, and will be deleted from the platform.

You will create one file for each audience you want to onboard. The file and field requirements are outlined below. Several file types are accepted. Please follow these guidelines closely to avoid errors.

• US records only
• Max file size < 20 GB
• File Types
  -    pipe-delimited (|) end in “.psv” extension
  -    comma-delimited (,) end in “.csv” extension
• Optional compression: Gzip (.gz) only
  -    When using gzip compression, you must ensure the extension .gz is added to the file name: MyAudience.psv.gz

Audience files consist of a header row containing column names, followed by rows of data that identify the audience members. See Section 6 for the complete list of allowed column names and descriptions for each. Supported columns are:

• PII (Name, Address, One or more email addresses – up to four)
• MAID
• IP address

To build an audience file,

1. Choose the columns that match the identifying information you have for your audience,
2. Write those columns into a header row in any order you choose, and then
3. Fill in the rest of the file with the identifying information rows, ensuring the data meets the formatting rules for each column.

Pipe (the “|” character) and comma (the “,” character) delimited files are accepted.

• Each file represents one complete audience.
• The first row of a file must be a header row containing column names.
• Column names may appear in any order.
• Subsequent rows will contain data records, each consisting of consumer identifiers for individuals in the audience.
• Empty or blank data values look like two consecutive delimiters (e.g. “,,” or “||”), or just one preceding or trailing delimiter at the beginning or end of the file.

Example of a comma-delimited file:

There are no special formatting rules for names and addresses, just

• Keep them under the maximum length listed in
• Only one address is allowed per row, and
• Follow the rule for quoting embedded delimiters.

Email addresses must be in the standard format local-part@domain.

You may include up to four plain text emails and four hashed emails of each type in one row. Each email must appear in a separate column and correspond to the column keyword.

Before hashing an email, you must normalize it by trimming leading and trailing white space and by changing all characters to lower case. See Hashing Data for more details.

Mobile Ad IDs come in two popular forms, IDFA (typically used with Apple devices) and GAID or Google Ad ID (previously known as AAID).

    Both IDs are 36 characters long and look like this:

    38400000-8cf0-11bd-b23e-10b96e40000d

Before hashing a MAID, you must normalize it by trimming leading and trailing white space and by changing all characters to lowercase. See Hashing Data for more details.

IP addresses may be provided in IPv4 or IPv6 format.

• IPv4 – Standard “dotted quad” notation. Example: 127.0.0.1
• IPv6 – For unhashed values, any standard IPv6 format is allowed.

Example: 2001:0db8:85a3:0000:0000:8a2e:0370:7334

Hashed IPv6 values

IPv6 values must be formatted prior to hashing.

To format prior to hashing, follow these steps:

1. Remove leading and trailing spaces and convert to lower case.
2. Remove the last four hextets.
3. Remove consecutive sequences of 0000 hextets.
4. Remove leading zeros from each hextet.
5. Append ::

Example

Original value: 2603:0000:0000:0DA0:7cd1:6410:25fa:7bf8
Formatted value: 2603::da0::

Let’s say the last name for a consumer has a value of Smi,th, maybe because someone typed their name in wrong. In a comma-delimited file, this value must be enclosed in double quotes: "Smi,th" and it would look like this:

Your source data may include hashed forms of MAID, IP address, and/or email address.

If you are hashing values yourself, you must transform values to lowercase and trim leading and trailing spaces before hashing. This applies to all hashed data (MAID, both types of IP address, and email address). IPv6 has some additional requirements outlined here.

For example, “  JSmith@Example.com ” becomes “jsmith@example.com” 

Put hashed data into its own column and prefix the column name with the hash used.  

Do not include raw and hashed versions of the same value in a row. For example, including both the email and the md5 hashed email is unnecessary. Instead, you should choose one and leave the other column blank.

Example: Do not include the md5_email in red below.

Specific file naming is required for the system to process correctly. Files should be named following this pattern.

• Angle brackets < > mean you must choose one of the options inside them
• Square brackets [ ] mean the contents are optional

clientname_audiencename.<psv or csv>[.gz]

1. The client name and audience name you choose
   -    Client name – add your brand/client name first, then an underscore (_)
   -    Audience name – add an audience name
2. A period (“.”)
3. The appropriate file extension depending on the delimiter used
4. A .gz if optional gzip compression is used

• Client name combined with audience names must be 100 characters long or less.
• Client names must start with a letter.
• Upper and lower-case letters are allowed but be aware that the final name in Experian’s activation platform will be changed to ALL CAPS.
• Underscores will replace any characters other than 0-9, a-z, and A-Z.

These characters will be included in the client and audience name as-is:

• 0-9
• A-Z
• a-z, although these will be capitalized
• underscore (_)

All other characters will be replaced by underscores.

Examples shown are in comma-delimited (.csv) format.

Example 1: Just email addresses

Let’s say your audience data contains only email addresses and for some people you have two emails. Your audience file would look like this:

Example 2: Name, address, and hashed MAID with embedded delimiter

Example 3: IP addresses

If you want to create your audience file using Excel, it’s easy to convert the Excel file to a comma-delimited file.

1. Open the Excel file that contains your audience data
2. Make sure to delete blank columns or rows that previously had data. Do not just delete the data that was in the cell; delete the entire column or row, or the conversion will not be successful
3. Go to File > Save As
4. Select the location you want to save the file to
5. Under Save as type or File Format, select:
   -    CSV (Comma Delimited)
   -    CSV UTF-8 (Comma delimited)
   -    Comma Separated Values
   All have a “.csv” extension.
6. Save the document

You can view your comma-delimited file by opening it with “TextEdit”.

We ask that you create a one-time test file once you have established your file transfer method and understand the file requirements.

Please create a test file that conforms to the specifications found in this guide. Prefix the file name with the word “Test”. For example, “Test ClientABC Audience_20250101.psv”.  

You can include real audience data in the test file, but keep in mind it will not be usable in any campaigns—this file is just to test the process and make sure everything’s working. 

Deliver the test file via your chosen method and inform your Experian representative via email.  

The Experian audience activation team will then review the file, help you correct any errors, and reach out to give you the approval to start sending real files for onboarding. 

Only include the columns for the identifiers you are submitting in the file. Your file cannot contain the same column name multiple times; they must be unique.