nScreenMedia OTT multiscreen media analysis

What’s working, and what’s not with OTT video experience

OTTtv WS Optimizing viewer experience panel

What makes a great online video experience? Here are three things that are helping to make it great, and three that could be making things worse.

At OTT TV World Summit in London last week I had the pleasure of moderating a panel entitled Attaining Optimization of Viewer Experience. I asked the three panelists to share what technique or approach they have seen that is helping to make the online video experience better, and what is not. Here’s what each had to say.

Techniques that are exceeding expectations

Adaptive bitrate streaming

Gary Hamer, SVP Sales and Business Development at SmartLabs, focused on one of the fundamental buildings blocks of the OTT video revolution:

“Adaptive bitrates have revolutionized viewing on multiscreen devices. We’ve certainly seen greater customers satisfaction with the quality of the video. It’s much better than just high, medium, and low, and you’re stuck with whichever one you chose.”

Virtual channels

Barney Withers Green, Sales Director, Media & Entertainment Solutions EMEA at Verizon Digital Media, zeroed in on the oldest form of video distribution, the linear channel. However, the way these channels are being put together online is undergoing a quiet revolution:

“The concept of virtual linear, or pop-up channels. Gone are days when you can just deliver a prescribed <programmed> channel. Whatever your content, you can create an engaging, customized channel. Our customizers have seen a huge amount of success with this.”

Data delivers hidden truths

Data is one area that is receiving a huge amount of focus in the media industry. Henrik Eklund, CEO at Newstag, is seeing a lot of user data since his news video service launched. Moreover, aspects of the aggregate data have surprised him:

“What surprised us the most is when we launched this concept of crowd curation and looked at the aggregate data, we assumed it would be really narrow and be about the Kardashians and Trump. But that’s not really the fact.”

Techniques that just don’t measure up

Recommendations

One of the areas where user data is applied most frequently is in the area of content recommendations. Mr. Hamer does not believe recommendations are necessarily helpful to everyone, especially pay TV operators.

“I’m going to be controversial and say recommendations. In a service provider environment, the experience is disappointing. They are not typically drawing from an infinite source of content. They have different relationships with different content providers, and the quality of the metadata is often different. The quality of the recommendations is disappointing.”

Mr. Eklund also thinks recommendations have been a disappointment. However, he thinks his company, Newstag, may have the answer:

“We have a really fast-moving environment, and 50,000 stories with all of them being consumed all the time. We have <fast-moving> data that would be interesting to translate into a long-form experience. If this data can be translated to my mood, I think they <recommendations> can be valid again.”

Balance between latency and buffering

One way to fix latency, a measure of how far a live stream is behind the original event, is to reduce the amount of video that is buffered by the client device. However, a smaller client buffer means the video playback is more vulnerable to freezes. Mr. Withers thinks many haven’t got the balance right yet:

“Latency is the big thing we are seeing lately. People are trying to get the most out of HLS, trying to balance buffer ratio and latency. There’s a lot of people that are trying to get the balance between the two and some of them aren’t getting it quite right.”

Why it matters

Not every idea is a good one when trying to improve the streaming media experience.

Adaptive bitrate streaming, virtual channels, and lots of user data seem to be working.

Recommendations and latency reduction techniques are not.

Facebooktwittergoogle_plusmailFacebooktwittergoogle_plusmail

One Comment

  1. Pingback: 24/i

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.