Key Concepts: Data Sources and Profiles
What are Data Sources ?
Watching That is essentially a data pipeline at it’s core. To make it work data sources need to be connected to your acount.
Data sources that are supported include:
Pixels - a pixel is the colloquial name given to a URL that can be called from any device to register an event and its state/context. In Watching That a Pixel can be deployed via an ad server as a 3rd party tracker, within native code as an analytics sink or called from any internet connected device. The firing of a pixel represents an event type (app load, ad impression, play request etc) that then powers the Metrics in Watching That. On the URL that is fired, key value pairs can be applied in the query string. These keys map to Dimensions in Watching That.
Log level data feeds - many technology systems like ad servers, SSPs, SSAI platforms etc provide log level feeds of their experience. For example Google Ad Manager provides Data Transfer Logs that is a complete and continuous list of all event data it has received and provided (ad request, code served, impression, ad complete etc). The rows of events map to Metrics and the columns map to Dimensions.
Spreadsheet/Excel reports - many businesses interact with summary reports such as daily reports that are presented in spreadsheets. Watching That can consume these reports automatically. The rows of events map to Metrics and the columns map to Dimensions.
System APIs - modern systems provide APIs that allow other system to interface with them without an human assistance. These APIs provide access to highly contextual information about their configuration and state at any given moment. Watching That can query system APIs as required. The results back from these call typically map to Derived Dimensions in the Watching That platform.
Tip: Watching That can collect from pretty much any data source that you can make available.
What are Profiles ?
A Profile is a predefined segmentation or bucket of data within a single source. Rather than treating a data source as one monolithic set of information, a Profile lets you organise the feed into meaningful top-level categories. This segmentation helps you tailor metrics, reporting, monitoring and user access to specific subsets of your data without clutter or overlap.
For example, if you provide an ad server log feed that contains all you ad inventory data you could create separate Profiles for:
Different Products / Apps
Offsite vs Onsite Inventory
Different Geographic Region
Watching That allows for Profiles to Rollup in the Reports Module because they support a hierarchical (parent/child) relationship.