Skip to end of metadata
Go to start of metadata

Auction Segment Feed

This feed is not available to Xandr Invest clients.

This is a log-level data feed that provides you with information on all eligible segments in an auction that your seller member has access to. The segments in this feed include pixel, query string, and 3rd-party shared segments.

The AppNexus internal name for this feed is auction_segment_feed.

Each feed is delivered hourly and will be retained for 7 days.  The feed will be logged when an auction is complete.

If you enable the auction segment feed, it will be in your feed list along with all your other log-level data feeds. See Log Level Data Feeds for more information (customer login required).

UNIX Epoch time

The date and time (e.g., 1526057561 which would need to be translated to Friday, May 11, 2018 4:52:41 PM (UTC)).

The unique ID of the auction.

The user ID.

The member ID of the seller.

A list of segments. Each segment consists of:

  • segment_id (int): The segment ID of the pixel, query string, or shared segment.
  • segment_value (int): A value that can be passed with the segment.
  • last_seen_min (int): The number of minutes since this user was last tagged with this segment. Given in Posix time (represented in minutes). Multiply this value by 60 to get the number of seconds since the epoch when this segment was updated.


The hashed version of the AppNexus 64-bit User ID which will we provided as a proxy in certain cases where AppNexus is unable to provide the real user_id_64. You will not be able to target users via their hashed user ID. However you can use this identifier to calculate unique users, user frequency, and user recency. An example is provided below:

If you receive the  user_id_64  version of this field, you will not receive the hashed version.

7partition_time_millisUNIX Epoch timeThe hourly partition any record row falls into, represented as a UNIX Epoch timestamp value (in milliseconds). Useful for defining hourly partitions when loading into a database or data warehouse. For the Avro format, this field is given the timestamp-millis logical type for native timestamp detection. For example,  1568077200000  can also be represented as  Tuesday, September 10, 2019 1 AM  (UTC).
  • No labels