If you collect and maintain segments of your own (i.e. for use in targeting), you can add these in AudienceReport, and this way measure on these parameters. In order to use your own data for Custom Segments in AudienceReport, you need to upload segment data to a dedicated S3 location (online data storage), from where it will be processed once per day.
Each segment can be made available to all AudienceReport customers, or it can be private and just opened up to individual companies as requested by the data provider(you).
The upload should take the form of one or more comma-separated text files (csv), which must conform to the format specified below. Furthermore, each segment you wish to provide needs a set of identifiers and labels, which will be used when reporting the results in AudienceReport. This setup step is done in cooperation with the AudienceReport helpdesk, who will ensure that the custom segments work as intended before they are made available.
Segment data file format
Files should be in plaintext, named according to Amazon S3 guidelines, with columns comma-separated, and with the first row of each file containing valid column headers. For example:
Column 1: keyspace
The first column in each row must contain the identifier for the keyspace, which is what cookie ID pool the user ID belongs to. For example, if the userid column contains Adform UUIDs, the corresponding keyspace column must contain "adform".
If you wish to provide custom segments using your proprietary identifiers, you will need to setup a cookie synchronization with AudienceReport beforehand, in order to identify your segments among the AudienceReport traffic. Once the cookie sync is in place, you will be assigned a keyspace identifier which you may use in your custom segment uploads.
Note: It is not currently possible to include rows with different keyspace identifiers in a single segment file. In case you are building segments from different sources, you should supply a separate file for each keyspace.
Column 2: userid
The second column must contain the user identifier belonging to the specified keyspace from column 1. This will usually reference a common third-party cookie pool, such as Adform UUIDs, or, in the case of segments based on UserReport survey questions, UserReport UIDs. As previously mentioned, it is also possible to use your own user/cookie/device identifiers, though this requires a cookie sync setup, and may result in significantly reduced sample sizes.
Column 3+: segment dimension columns
The remaining columns specify the segment dimensions contained in the file. You may upload a single segment dimension column per file, or choose to include multiple segment dimensions in one file, by simply appending more columns.
A 'segment dimension' corresponds to a single multiple-choice question, such as "Are you in the market for a new car?", "Which party did you vote for in the previous election?", or "How likely are you to recommend Cyberdyne Industries to a friend or colleague on a scale of 1-10?".
The first row must specify the segment ID for the column, in order for AudienceReport to digest the segment data correctly. For example, in the above example we have two segment dimensions, identified by their segment IDs:
The numbers in each row are 'segment outcomes', corresponding to the possible answers to the multiple-choice question. Segment outcomes must be positive integers, and the outcomes will be displayed in reports in the order corresponding to the integer order. For the above segment dimensions, we might have the following possible segment outcomes:
Dimension and outcome labels for segments are chosen by the data supplier and entered into the Custom Segments system. Note that the maximum number of distinct segment outcomes for a single dimension is 6.
You can reach out to email@example.com or your usual contact at AudienceProject to learn more about how to get started with custom segments.