Getting started with the Kaltura Analytics API
Kaltura Analytics Overview
Kaltura Analytics helps you get precise, actionable insights into various aspects of your business by answering a wide variety of questions, such as:
User Behavior Questions:
- How are my videos doing?
- Who is watching them?
- On what devices?
- What is popular now?
- What is the quality of service my users are experiencing?
- Are there any glitches?
- How can I improve my quality of service?
- What is my system usage?
- Am I close to my allowance?
- How do I charge back for usage expenses?
- How do I optimize my budget?
With Kaltura Analytics you can collect, present and analyze data about system usage, viewing behavior, and any other type of user interaction with the Kaltura solution.
The data in Kaltura Analytics can be divided into roughly two groups: player events and other data.
As a video is played, the Kaltura player sends player events related to viewer behavior (such as “Plays”, “Played 25%”, “Stopped”, etc.) to Kaltura Analytics. If you are using Kaltura Player versions XXX or higher, you will get the support for these player events analytics out-of-the-box. If you are using a non-Kaltura player, you need to make sure all of the relevant events are reported correctly using Kaltura Event Tracking API. To learn how to configure your non-Kaltura player to report relevant events, click here.
Every player event is related to a specific video asset (a VOD or a live stream) and is triggered by a specific user. Every event has certain pre-defined characteristics (such as time, device, geography, etc.) that are called dimensions.
Let’s take a look at an example.
Suppose that at 1 pm EST a user in California clicked the “play” button on her iPhone device (the device has an IOS 9.1 installed on it) and begins to watch video asset #1234.
In this case the recorded event is “play” and the dimensions are:
You can read more about currently supported dimensions here.
These events provide a basis for measuring metrics, which are later presented in Analytics reports, accessible through the Data Retrieval API. Metrics refer to the data that Kaltura Analytics measures, which are quantitative measurements. These measurements can assess user actions (such as user activities, plays, comments, drop offs, etc.) or resource consumption (such as bandwidth, storage, transcoding, etc.).
You can read more about the currently supported metrics here.
Metrics are stored and presented in an aggregated form. Aggregations are a way to store the huge amount of data in a summarized, pre-computed format that allows fast rendering of reports. The aggregations are designed with your business objectives in mind and anticipate questions you might ask. For example, Kaltura anticipates that you might ask the question “what are the most popular viewing devices in various states?”, so we created an aggregation of device X geography for the play requested metric.
You can read more about currently supported aggregations here.
Every report in Analytics is made up of Dimensions, Metrics and Filters. Dimensions and Metrics are explained above. Filters determine how the data will be filtered. For example, instead of returning all user activity for a particular category, a report can return only data for a certain country.
What to read next:
Dimensions dictionary: Features a list of the dimenstions and their meaning within the Kaltura Analytics project.
Metrics dictionary: Features a list of the metrics and their meaning within the Kaltura Analytics project.
Player events guide: https://kaltura.atlassian.net/wiki/display/KANAL/kAnalony+Event+Tracking+API - NEED A DIFFERENT LINK - THIS IS THE SAME AS THE EVENT TRACKING API LINK!!!
Supported aggregations: Features an overview and actual aggregations implemented for the following supported aggregations:
Kaltura Analytics User Guide:
- Event Tracking
- With Kaltura Player
- With non-Kaltura Player
- Data Retrieval