Follow

General Information

The StrikeAd generic Data Management Platform (DMP) integration framework is designed for partners and clients who want to expose their own data set for targeting in ad campaigns.

  • Segments are stored in the Strikead targeting database and become available for targeting in the platform.
  • Each partner that is added is represented as a separate targeting option at line item level in the Fusion user interface/API.
  • Segment permissions can be set for each Fusion user base or for each Fusion agency base.
  • A CPM cost model can be applied to the usage of the segments.

File Formats

Master File

Currently the DMP framework only supports appending to the master file (adding new segments and changing the name of an existing segment). The master file includes a list of segments and metadata, including taxonomy. The master file is a csv file and uses the formats described in the following table.

CSV file columns names Description Example
id Segment ID (integer) 350037
parent_id Parent segment ID (integer). If the segment does not have a parent, an empty value is allowed. A parent segment must be defined before its child segments. 24148
name Name of the segment that will be present on the Fusion user interface/API for targeting. 150 symbols long. Segment 1
size Estimated size of the IDs in the segment (integer). 141651840
price Cost per CPM in USD. 2.15

Save each new master file with the incremented version number in the object path, for example,

{basePath}/masterfile/masterfile_{version}.csv

Where version is the incremented integer.

Tip: It is possible to use a reversed date+time, such as: 201606121234, or timestamps.

The DMP framework maintains information about the current applied master file version. Periodically, the DMP framework checks for a new master file (approximately every 5 minutes). If a new master file is found it is applied, and feedback is provided through email.

Example
{basePath}/masterfile/masterfile_1.csv
{basePath}/masterfile/masterfile_2.csv

A sample file is attached.

Scheduled Master File Updates

The DMP framework checks the master file location hourly. If the master file has changed, the DMP framework uses the new file. Parent/child configuration for segments cannot be changed in the new file. If teh parent/child configurations is changed then the appropriate segments are ignored, while all other segments are replaced.

Segment Content Files

The segment content files contain the actual content of the segment. A day before the campaign starts, the DMP framework pushes the files to the database for targeting in campaigns. The data is purged from the database if there is no targeting for this segment in the Fusion. DMP framework.

File Formats

Segment content files can be in any of the following formats:

  • .csv (comma-separated)
  • .csv.gz
  • .csv.gzip

It is preferable to provide the segment content files using .gz or ..gzip files, where the content is zipped, as this saves space and throughput.

Uploaf each segment as a separate file. If needed, you can split a segment into multiple files. 

The segment files can contain mixed device ID types, and the segment name should contain the segment ID (the same as in the master file).

Naming Conventions

Use the following naming conventions for the segment file:

{basePath}/segment/{segmentId}/{segmentId}-{reversedDate}-{time}.csv
{basePath}/segment/{segmentId}/{segmentId}-{reversedDate}-{time}.csv.gz

Where:

  • The segmentid is used as a folder name and in the file name
  • The Date must be in the format YYYYMMDD to allow for alphabetic sorting
  • The time must be in the format HHmm

Example:

{basePath}/segment/53256/53256-20151224-2145.csv 
{basePath}/segment/53256/53256-20161224-2145.csv.gz
Supported Device ID Types
  • idfa
  • idfamd5
  • idfasha1
  • did
  • dpid
  • didmd5
  • dpidmd5
  • didsha1
  • dpidsha1
  • raw - for unknown device ID types
File Content

The following is an example of the structure of a segment file.

#version: 1 
idType,deviceId
<idfa,foo
idfasha1,0beec7b5ea3f0fdbc95d0dd47f3c5bc275da8a33
idfamd5,acbd18db4cc2f85cedef654fccc4a4d8
...

Important:The version number must be set to 1.

Sample File
Schedule for Pushing Segments

A day before the campaign is due to start, the segments are automatically pushed to the StrikeAd database. The segments are available for targeting in the StrikeAd database for the flight time of all campaigns that include targeting to these segments. If there are no campaigns that include targeting to the segments, they are automatically removed from the database.

Increments Support

Increment support enables the DMP to provide frequent updates (for example, daily, by saving the the increment files in separate folders.

Use the following conventions for the increment file:

{basePath}/{segmentId}/increments/{segmentId}-{reversedDate}-{time}-increment.csv

Where:

    • The segmentid is duplicated as a folder name and in the file name
    • The Date must be in the format YYYYMMDD to allow for alphabetic sorting
    • The time must be in the format HHmm

Resolved as follows:

    1. Find latest snapshot.
    2. Find all increments after this snapshot. If the snapshot does not exist, take all increments.

Example:

{basePath}/segment/53256/53256-20151224-2145.csv 

File Storage

The S3 bucket is provided by StrikeAd or by the partner.

Reporting

StrikeAd enables you to generate reports that drill down into segements. The feed of the data can be published to the S3 bucket.

Feedback

For all actions such as checking for a new master file, or uploading a segment to the DMP framework, send an email of the DMP action log.

Was this article helpful?
0 out of 0 found this helpful
Have more questions? Submit a request

Comments

Powered by Zendesk