How should an SSP receive the Anonymised audiences
The Anonymised ID Less segments are passed to SSPs using the following header bidding solutions. For more information, refer to the relevant documentation:
Header Bidding Functionality
Both header bidding solutions work similarly; in that they:
Check that they and the audience vendor have consent to read and write from the browser's local storage
Read the Anonymised segment array from the browser’s local storage
Format the segment array into the expected object within the ortb2.x bid request
Amend the ortb2.x bid request with the formatted object
Ensure the amended bid request is sent to the SSPs bid adaptors as configured
Expected ortb2.x object and format
As per Prebid’s recommendations for handling First-Party Data, the Anonymised segments will be formatted and appended to the ortb2.x bid request as follows:
...
},
"user":{
...
"data":[
{
"name":"anonymised.io",
"segment":[
{
"id":"18112"
},
...
{
"id":"default"
}
],
"ext":{
"segtax":1000
}
}
]
}
}
...
Where:
The data extension object is used to pass our default segment taxonomy (IAB segtax) value.
You will note that our use of the segment taxonomy only identifies the segment key-value pairs as belonging to Anonymised and does not classify the segment values specifically.
This is intentional and part of our k-anonymity practices.
Exception to the ortb2.x object and format
Xandr (bid adaptor name: appnexus) does not ingest such data as segments instead it considers such data as keywords. Thus the ortb2.x.user.keywords object is used.
The keywords can be configured at auction or impression/adslot level but in all cases, the following format is expected in the ortb2.x bid request:
...
},
"user":{
"keywords":"perid=18112,..."
"data":[
{
"name":"anonymised.io",
"ext":{
"segtax":1000
}
}
]
}
}
...
Where:
We still expect to see the segtax value 1000 being passed in the data extension object.