Streaming Video Analytics API and Developer Tools

Looking for a streaming video analytics API and other developer tools related to video metrics?

IBM has expanded the analytics options available to IBM video streaming and enterprise video streaming users. This includes a set of APIs that allows developers to take metrics related to their video content and infuse them into other applications. These available analytics can be used to track content broadly, seeing metrics on all content. However, they can also be used to track individual usage, associated with an email address gathered either after a user authenticates for restricted internal content or fills out a registration gate for video marketing content.

This article covers what type of data you can get through using the streaming video analytics API, where to find them and API use cases. Some of these are focused on the idea that they will be used for video marketing, while others on the very different use case of internal comms. For the latter, also be sure to download our Using Video for Internal Corporate Communications, Training & Compliance white paper as well.

Building a streaming video analytics dashboard

In its own dashboard, IBM Watson Media provides a range of real-time and historical analytics related to video content.

Now IBM is introducing APIs to help bring some of this analytics data into other applications, such as custom dashboards, custom interfaces and more. This includes the ability to generate data for a specific time period, such as for the current month, or track all unique viewer. Furthermore, this data can be generated for all video assets, used to see for example the overall engagement on video content at a given moment, or to narrow down on a specific piece of content as well.

These analytics can be broad, but also have the capability to dive deeper, looking at a viewer on an individual level.

Viewer: individual viewer tracking

One of the powerful aspects of the analytics offering is the ability to track individual users. Depending on your use case, this can be done through associating that viewer with an email address. Now there are two ways to gather this information and start tracking them. The first is for restricted, internal facing content through IBM’s enterprise video offering. This requires users to authenticate through a SAML based SSO (Single Sign On) method. As a result of this authentication, the viewer can now be associated by their email address. Alternatively, content owners can also elect to place their content behind a registration gate, which can ask things such as job title or company from those who register. One of the required fields, though, is the viewer’s email address, which after being entered also allows them to be tracked.

Once an email is in the system, the APIs can be used to track these individuals. Data available through the APIs, on a per email basis, includes:

Last activity
Number of videos watched

In addition, there is also a lot of detail to be found looking at the view data.

View: session data

The view parameter basically builds a profile around a specific viewing session. This enables rich details on the viewer, both about the device used and demographic information like where they are coming from. In total, it includes:

Geographic location (city and country)
Content type (live or recorded)
Access period (start and finish times)
Amount watched (point started and point stopped in the content)
How content was watched (mobile or not)

For reference, the geographic data is pulled based on the IP address used at the time the viewer started to watch. Now the reliability of this varies based on how precise it gets. Its been reported that at the country level the accuracy is very high on average, about 95-99% accurate. It gets less accurate at the state (55-80% accurate) and city level (50-75% accurate).

Check here for more information on how geolocation through using the IP address works.

Using the APIs

Developers can start using this Analytics API Documentation to begin integrating metrics related to their video content into other applications. This includes instructions on how to authenticate, to authorize their use in other applications, and a rundown of various functions from getting viewership data to being able to drill down and get information on an individual viewing session.

As the APIs are expanded, the documentation link above will reflect the latest version.

Analytics API use cases

There are a variety of different uses for the streaming video analytics APIs, both on internal and external setups. Below are a few that can be achieved through data pulled through the APIs from the available analytics.

Compliance checking for internal comms
Created some mandatory training? Content creators can track who has watched video assets along with the amount watched, given via a specific start time and finish time for their viewing session. From this, a list of viewers can be gathered and then cross checked to see who might be missing or who didn’t watch the content for an adequate length.
Gauging content performance
Across the board, a major benefit of metrics is gauging the success of created collateral. Data on visitors, including raw number of viewers on content or how engaged specific viewers were, go a long way from bridging the gap between thinking content is successful and actually determining it is. Bringing this data into other dashboards and interfaces can help put this important information in more places, enabling more people to gauge how video assets are performing.
Custom messaging for viewer marketing
Viewer and view session data can provide powerful insight into a user. Information like geographic data and how they watched can be valuable. For example, maybe you have services that have different offerings available over laptops compared to those on mobile. As a result, discovering that they are watching from a mobile could influence the type of message you send them, with links to more relevant content for them. Additionally or alternatively, where the viewer is coming from could also be used to customize messaging. That said, keep in mind that while the country reported has a high accuracy level, things like the city have a lower confidence level.
Qualifying leads for video marketing
Certain characteristics could indicate that a video marketing generated lead is more qualified than others, and these characteristics can be gathered through analytics related to their viewing. For example if the viewer watched a large number of videos that could be seen as a positive. Recent activity could also be an encouraging sign, and perhaps indicative of a lead ready for a sales follow up. If your product is only available in certain markets, geographic data could also help determine how qualified a lead might be.

Summary

Getting metric insights about how video content is performing can be key. The ability to do so from custom applications, like dashboards or interfaces that you manage directly, can not only be convenient but also can broaden who gets access to this information. As it pertains to aspects like training, usage data can be infused in other services, sometimes boiling it down to just the needed information. Similarly, on the topic of lead qualification, it can also be focused on pertinent information that could help to validate how qualified a prospect is.

These APIs are available now to IBM video streaming and enterprise video streaming users across a variety of use cases. If you are more interested in using video for marketing purposes, also be sure to download our How to Increase Conversions with Online Video white paper as well.