Length, Size, Going to Completion Matter….for Online Video Viewing by Device

Length, Size, Going to Completion Matter….for Online Video Viewing by Device

Share on

It’s a week of provocative openings for me I guess and no matter what anyone says, length and size do matter… especially in regards to online video. Perverts. Looking back at the Ooyala Q2 2012 Global Video Index, I found some further interesting bits and pieces that I thought might help some of you with your online video campaigns for marketing, branded content and original web series.


From the Global Video Index:

Ooyala measures the anonymized viewing habits of nearly 200 million unique viewers in more than 130 countries every month. We process billions of video analytics events per day, and use our market-leading video analytics to help media companies and consumer brands grow their audiences and earn more money with mobile, multi-screen broadcasting.

So while not a completely accurate representation of the entire global online video consumption, it’s a pretty good snapshot of viewing habits.

The Long and Short of IT

The “IT” being online video of course. You’ll remember that back in the Q1 2012 Global Video Index, Ooyala found a disparity in the time of digital video watched based on the device. This quarter, that disparity continued to evolve. Longer video is taking over more time on all but the smallest screen, mobile. Even Connected TV and game consoles saw a noticeable shift toward longer content, 10 minutes plus.

This pretty much means that you can now start making longer content in general because people are becoming more accustomed to viewing it across all screens. Even mobile, which I said was sort of backsliding, is still at just under 50% in terms of 10+ minutes and if you add in 6-10 minute videos, it’s more than 50%. The long of it is:

Videos of 6 minutes or more are very popular on all screens.

On the short end of the spectrum, videos less than a minute in length, were not at all popular. Just this week, I wanted to watch a 57 second video and lo and behold, there was a 30 second ad in front of it! Well, you know exactly what I did… I went to go find one without that ad. Sure, it probably took me more than 30 seconds to do it, but it’s a vital lesson in consumer behavior.

The videos that were still, fairly popular across all but the largest screen, were in the 1-6 minute range. In fact, they account for almost 50% of all videos on mobile and topped 25% on desktop and tablets.

I slightly modified this Ooyala chart below. I wanted to put the results in order in order to make a point so I swapped some of the locations.

The platform that breaks the generalization here is tablets. I would say generally:

The bigger the screen, the longer the content.

But that’s not quite right is it? I mean, tablets sort of break that statement. I have a theory about that, you see, tablets are more portable, can be stood up, propped up or mounted. Meanwhile, a desktop is moored to the boat anchor of a desk and laptops generally need a nice flat surface, and aren’t quite as portable. So what I think is happening, is that since the tablet is more convenient, it’s being used in more locations, more often which is why it gets more long-form video, based on percentage of viewing.

For comparison, here is the Q1 2012 chart:

Comparing the two charts it seems to me that tablet viewing is massively evolving this year and that’s causing some minor evolution in desktop and mobile. Connected TV is exactly where I’d expect it to be, because it’s a TV screen.

The Evolution of Digital Video Viewing by Device

In fact, Ooyala even stated something along those lines in the GVI:

The share of time spent watching video on mobile phones and tablets continues to grow quickly. The overall share of mobile video plays jumped 36% last quarter, while tablet video’s share of time watched increased by 72%. Together, the overall share of time
spent watching video on mobile devices and tablets increased 64% in Q2 2012.

But how much time does that total up? They say, on average, globally, about 5%.

In the U.S. alone the combined mobile and tablet viewing is just over 5% of all online video viewing. Still not a massive portion of viewing, but it’s growing and will probably continue as new tablets, new smartphones (iPhone 5, Windows 8, etc) come along and make it all the more enticing. In fact, I have to believe that this viewing is directly driven by the launch of newer, faster, smarter phones and tablets. So each new iteration of devices, which lowers the price of the previous iteration, means more potential for more online video viewing.

Going Long!

Length per video view varied across the quarter with them generally being lower in April and rising through June but dipping at the end of the month. Mobile actually ended lower than it started, while desktop and tablet rose and connected TV stayed about the same. They all exhibited fluctuations though showing that video viewing might still have some seasonality to it. I expect the next quarter to be weird, because of the Summer Olympics, which will certainly provide a massive uptick in online video viewing.

If you’re planning on putting out some video, and want to make it screen specific, you might want to consider adjusting the times. Obviously, you can’t always do that will totally scripted content. But you might be able to do it and put in a call-to-action on the smaller platforms that says “for the longer version, head to XYZ with your desktop browser or connected TV.”

Check out the time per play by device (below). If you can manage to shave off some seconds, you’ve got a better chance of the viewer sticking around for the whole thing on desktop and mobile. Tablet viewers are already going long with average viewing per video at around 3.5 minutes. Of course, no one goes long like the connected TV crowd, do they?

Going to Completion

Another key thing to think about, when tweaking that video content for each device, is making it to completion. The fact of the matter is that, getting there, is challenging. Even on tablets, users are only passing the three-quarter mark around 55% of the time. That’s not terrible, but when you look at total completion rates, there’s more drop off. Desktop and tablet viewers manage that just about 40% of the time. Mobile viewers show far less staying power and are closer to just around one-in-three times while connected TV users are closer to 50/50.

What you should take away from all that, is this:

Call-to-actions need to be earlier, rather than later most of the time.

If you’re putting some kind of call-to-action (CTA) in your video, it might be best to do it early. Better yet, you might do it early, in a non-intrusive way, and then reiterate that call near the end of the video. That will result in more people actually seeing the call-to-action. Check out the graphs below for proof.

The engagement numbers aren’t bad, and they’re definitely a helpful metric to look at because if you were just placing your CTA at the end of the videos and wondering why conversions were extremely low, you can see why. If you put the CTA in at the 50% point it could have a positive impact on conversions, especially on tablets and connected TVs. It’s a hypothetical conjecture on my part based solely on the fact that more people would see it. Based on that same metric, if you put it at the 25% mark, the benefit could be massive.

Check out the full Ooyala Q2 2012 Global Video Index

Share on


September 2018

View Leaderboards

Most Popular

Read More Insights

mobile video platforms

Skippable Mobile Video Ad Completion Rates Low, CTR Near Non-skippable

The Mobile Marketing Association has put out its first report on mobile video benchmarks. As a first go round it's not bad and offers some interesting insights, while lacking some more specific information. Still, it does a decent enough job of looking at CTR and completion rates for skippable and non-skippable mobile video ads in three major formats, linear video, interstitial and value exchange (incentivized).

Read More Arrow pointing right
© 2018 Tubular Insights & Tubular Labs, Inc.