Looking for HughesNet U.S.? Click here.


How Video Analytics Improve Enterprise Learning

How Video Analytics Improve Enterprise Learning

With video-centric learning becoming commonplace in enterprises of all sizes, learning organizations are discovering the value of video analytics in improving efficiency and effectiveness across their businesses.

Traditionally, companies have relied on reporting from their learning management systems (LMS) to glean insights on learning performance.  Aside from tracking compliance, the LMS offers good data o both macro-level trends and individual performance.

Video analytics compliment the LMS, offering deeper understanding into a learning community’s behavior.  This data provides actionable intelligence that directly leads to better business outcomes.

Here are a few examples of the power of video analytics in learning:

The Audience Retention Graph

The audience retention graph best exemplifies the insights that video analytics offer learning organizations.  This graph depicts the total audience for a given piece of content over time, giving visual representation for the “fall off” of viewers over the length of the program.

When analyzing Audience Retention, it is best to break this graph into three elements, the head (beginning), the body (middle), and the tail (end).

The head refers to the first two minutes - or two percent - of the video, whichever is less.  There are typically two reasons why viewers exit a video in its opening moments.  The first is if the viewer selected the “wrong” video.  This is important data for a learning professional, allowing them assess the title, taxonomy, and search criteria to determine how the system can be improved to more readily connect viewers with desired content.

The second - and more disturbing - reason for early drop-off is that the viewer does not perceive value in the content.  This alerts the learning professional to a potential problem with the content.  Is it too long-winded? Too ambiguous?  Not optimized for the viewing device?

Similarly, drop off in the body (middle 96%) of a piece of learning content is a strong indicator of a problem with the content presentation.  The audience retention graph provides insight into potential causes of the issue.  A sharp and sudden drop often denotes a point of large-scale audience disengagement.  This is typically associated with a segue to a less popular topic, or a learning exercise that does not resonate with viewers.

Another reason for audience drop-off in the body of a piece of content is the attention span of the viewer.  Enterprises need to train a broad spectrum of learners.  Associates in some job functions have shorter attention spans than those in others.  It is important for instructional designers to consider this when producing content.  Video analytics allow learning organizations to study a specific job function, determine its attention span, and design content accordingly.

A loss of audience retention in the tail (final 2%) of a video can be particularly troublesome for learning organizations looking to gain data through surveys and assessments at the end of the video.  There are content strategies - such as avoiding long recaps, or the use of terms like summary and conclusion - that help keep viewers engaged.

Polling Analytics

A common instructional design strategy is to interlace polling questions throughout a training course. This technique reinforces key learning objectives, resulting in heightened knowledge retention.

Analytics based on responses to these polling questions offer value by providing insight into the quality of the instructional design and by identifying disengaged, or struggling, learners.

Aggregated polling results can be analyzed to compile the percentage of correct responses to a given question.  Questions with fewer correct responses can be identified and studied by the instructional designer.  Quite often this identifies opportunities to improve the clarity of the lesson, or to improve the question itself.

For live, interactive distance learning (IDL) courses, an instructor can be alerted in real-time should a student repeatedly respond incorrectly.  These notifications allow the struggling learner to be directly engaged for one-on-one remediation, improving that learner’s performance without negatively impacting the audience at large.

Another analytics capability of browser-based interfaces is the measurement of the time-delta between when a question is issued and when a learner responds.  This measurement provides live IDL courses with awareness of the questions that “hang students up.”  It also can be used to identify a disengaged or distracted student.

Search Terms and Viewing Behaviors

class="Body" style="text-align:justify">Analytics derived from search terms and viewer behavior provide additional value to learning organizations.


Search terms help learning organizations better understand the relevance of their content.  As corporate learning moves to less formal, more just-in-time applications, such as Performance Support and Micro-Learning, this data helps drive successful content strategies by offering the “voice of the learner” in their moment of need.

Data derived from viewer behavior - including content type, audience (job function), time of day, and viewing device provide essential data to instructional designers, resulting in better designed content and results.


As Lord Kelvin famously stated, “If you can not measure it, you can not improve it.”

Video analytics not only provide the measurement, but actionable intelligence based on those measurements.  This helps learning organizations improve their learning content and better engage with their students.  This translates into better performance across organizations, with compelling return on investment, measurable in items such as employee retention and operational efficiency.

As businesses becomes increasingly focused on outcomes, video analytics helps learning organizations drive - and measure - those outcomes.