LongTail Community Blog

Mobile platforms for video: understanding where and when

Share

Here at LongTail, we have a unique opportunity to observe and understand how video is being watched over the Internet. Our popular video player (JW Player) and video hosting platform (Bits on the Run) provide us with insight on video and device viewing habits. With the increasing popularity of tablets, we decided to have a closer look at just how this new class of devices account for video consumption across the Internet.

Not only are tablets emerging as a new class of devices in their own right, shipment numbers indicate that they might even overtake personal computers in the near future. In Q3 of 2011, Apple shipped 15.4 million iPad tablets, more than the total volume of personal computers sold by each of PC heavyweights HP, Dell and Lenovo. More recently, in Q2 of 2012, Apple shipped 17 million iPad tablets globally out of 24.9 million tablets in total (Android tablets accounted for 7.3 million of those).

So how are tablet users watching video? Which are the most popular tablets used for watching video? And at what times of the day or week is video playback peaking in tablets?

This blog post shares some of the findings we find most intriguing. Note that the tablet market - like any other emerging technology - changes rapidly (we have a note in this post about how the Nexus 7 is already shaking up the low-end tablet market). So stay tuned for updates!

iPad Dominance

Let’s face it: the iPad is ubiquitous. Currently, no other tablet even comes close in terms of challenging the iPad in its dominance of the tablet market, especially in terms of watching video. (The data in this section is for video traffic on our Bits on the Run hosting service, from September 2011 to May 2012.)


Tablet Market Share
Figure 1: Marketshare of various tablets (Sept. 2011 - May 2012).

It gets even more interesting if we break it down over U.S. and non-U.S. (international) markets.


international
Figure 2: Marketshare by Geo (Sept. 2011 - May 2012).

The iPad marketshare in non-U.S. markets is a whopping 95% for video streams, which makes it tower above all other vendors. In the United States, the Kindle Fire* seems to be competing for some of that market share (~5%), but only incrementally. As of now, it is really the only visible competitor to the iPad on the market.

*Note that the Kindle Fire is not being sold outside the U.S., this may mean that other tablet makers are having a hard time creating a solid competitor to the iPad.

Tablet Growth: Usage over Weekdays and Weekends

Tablet usage (in addition to smartphone usage) seems to be increasing over time. In our analysis of the overall growth of smartphone and tablet usage, we found that smartphones and tablets together account for over 10% of video consumption usage, and that their usage has been growing steadily over the past two years. (This section uses data from the free version of our popular JW Player, which is widely used on websites around the world.)


Smartphone Growth
Figure 3: Growth of smartphone and tablet usage in video (July 2010 - Feb 2012).

We also analyzed how people use their tablets and smartphones to watch video, over time of the day and day of the week. This data was measured over a period of two months, January and February 2012. (For hour-of-the-day calculations, we normalize the time of day from our server logs by adjusting the hour to the user's timezone.)


Usage by device
Figure 4: Device usage by time of day. Light blue indicates late night / early morning hours, pink indicates morning hours, yellow is afternoon / early evening, and green is late evening / night.

What we found was, over the course of a day, tablet users seem to predominantly use their tablets to watch video during the evening hours (after 6 pm). In contrast, connected device users (e.g. PlayStation) seem to prefer to use their device in the early hours of the morning, and so do smartphone users to some extent. Desktop users predominantly use their devices during the afternoon.

But over the course of the week, the variance of video consumption by various devices is smaller. That being said, phone and tablet usage are slightly higher over the weekends compared with desktop users.


Usage by day of week
Figure 5: Device usage by day of week. Light green is Monday to Thursday, dark green is Friday, and blue is weekends.

Android Tablet Share

Not many surprises thus far - the data presented above matches what we intuitively knew about tablet market share and usage. Next, we explored a currently growing area in industry: the Android tablet market share. Let’s see just how the Android tablets stack up in terms of popularity.

We counted at least 1,800 distinct models of Android over the period covered in this analysis. While these 1,800 models are distinct model numbers, they appear not to represent distinct devices. For instance, the Samsung Galaxy Tab tablet actually has different model numbers for each of its various screen size offerings (7 inch, 8.9 inch, and 10 inch versions). In addition, each screen size model might have different model versions for its different feature sets. For example, the Samsung Galaxy Tab 10.1 WiFi only tablet has a model name GT-P7510, while the Galaxy Tab 10.1 WiFi and 3G model has a model name GT-P7500. Samsung also seems to have different model names for the exact same device in different international locations, complicating things a bit more.

Our analysis of the Android-specific tablet market verifies that the Kindle Fire currently dominates the Android segment.


Android device marketshare
Figure 6: Market share of Android tables (Sept 2011 - May 2012).

As seen in the image above, the Kindle Fire accounts for the single largest Android tablet consuming video, accounting for 40.8% of all Android tablet video impressions on our video hosting internationally (the above graph is from November 17, 2011, when the Kindle Fire launched, up to early May). For this time period, we found the Kindle Fire to be the largest Android device consuming video across the entire Android phone and tablet range (though only tablets are represented in these graphs). These numbers seem to be corroborated by Flurry’s analytics data.

However, there is one interesting phenomenon that we observe. While the Kindle Fire has shown spectacular growth over the last few months, it seems to have peaked shortly after the winter holiday season (Dec 2011 - Jan 2012), and seems to be more recently declining. This is along the lines of what others have found.


Android table growth
Figure 7: Growth of Android Tables (Sept 2011 - May 2012). Solid black line represents non-specified Android tablets.

Interestingly, there seems to be no massive change in the growth of other Android tablets - even the Samsung Galaxy Tab 10.1 is showing only moderate signs of growth, indicating that Android tablets still have a long way to go in terms of challenging the iPad’s dominance. (The dark black in the graph that is second to the Kindle Fire represents all other Android devices not listed individually.)

The share of Android version numbers, across tablets, is shown below. You will see that the usage of Android 2.2 (Froyo) is dropping steadily, and Android 2.3 (GingerBread) has the largest marketshare, even among tablets. Android 2.3 numbers are actually boosted by the Kindle Fire, which runs on Android 2.3.4. In the meantime, usage of Android 3.x (Honeycomb) seems to be at a standstill, but usage of Android 4.x versions is growing steadily. (The drop in Android 2.2 in our graph in February 2012 is due to our dropping Flash support for Android at that time, which resulted in those users registering as downloads rather than streams on our system).

Wait, Android 2.2 tablets, you say? That’s what we asked ourselves too - until we found out that the majority of Android tablets sporting “Froyo” were either Dell’s 7-inch Streak and the Toshiba Folio, which accounted for about 60% and 30% of the Android 2.2 tablets that we saw in our measurements.


Market share of various Android versions
Figure 8: The marketshare of various Android versions, for tablets only (Sept. 2011 - May 2012). Note: Embed support was dropped for Android 2.2 in Feb. 2012.

At this point, you may be asking: "Hey, what about the Nexus 7 launch? How does that alter the Android tablet marketplace?" We have been asking ourselves the same question! The Nexus 7 is still in its early days - after all, it was only announced at the end of June, with the device reaching consumers in late July 2012. But, we have a sneak peek at the latest numbers that help answer the question of whether users are buying the Nexus 7 and using it to watch their videos...


Kindle Nexus Galaxy Tablet
Figure 9: A comparison of video streams from Amazon's Kindle Fire, Samsung's Galaxy Tab (all models and 10.1 specifically), and Google's Nexus 7 tablet that was just released.

... and the answer is, yes, they are! In fact, from our measurements, video streams played on Nexus 7 tablets are now already at about 35% of Kindle Fire streams - and that is only a few days after launch. Even more interestingly, when we include Samsung's Galaxy Tab in our measurements, we see that video streams from all models of the Galaxy Tab are starting to exceed those from the Kindle Fire for the first time since the Fire was launched. (Streams played on the Galaxy Tab 10.1 accounts for about 60% of all Galaxy Tab streams.)

What the Future Holds

As we noted at the beginning of the article, the tablet industry is still an emerging market, and we expect very interesting changes in tablet market share as well as video viewing habits as the devices and the market itself matures. As we just saw, Google's release of a low-priced Nexus 7 tablet is already having a large impact in the Android tablet market, though its too early to tell whether it will emerge as a competitor to the iPad.

There are some trends in the near future that will be interesting to watch: how will the upcoming launch of newer Android 4.1 (Jelly Bean) and Windows 8 tablets later this year shift the tablet industry and video viewing habits? Will Apple be able to continue its massive dominance of the tablet market (and hence video viewing on tablets) through new versions of the iPad?

These questions will be answered by the market and consumers as these new devices roll out. And when they do, LongTail Video will be tracking these trends.

Posted by Suman Srinivasan on August 01, 2012

Comments

As someone who loves watching the Apple V Android market, I loved this post Suman.

1. Did you see how Fig. 4 - closely reflects the hourly electricity load (demand) profile within a region. Consider the California Independent System Operator (CAISO) for example, http://www.caiso.com/Pages/TodaysOutlook.aspx#SupplyandDemand
The load profile is almost the same as when people watch videos - I thought that was very interesting, even though these devices dont contribue much to load profiles. In fact its merely co-incidental.

2. Its no surprise that we watch more videos during the week ends, but its great to see that as a graph. Loved that one.

3. People like reading more books during the holidays, and hence they buy more Kindles - probably, a reasonable explanation.

3. Kids with no morning jobs play - PS in the morning, probably another reasonable explanation.

4. Lets consider a hypothetical scenario - if Apple vanishes from the market tomm, then how long would it take for that fraction to come down from 91% to 50% ? Now, if Apple keeps pushing its boundaries, its going to take a long (er) time - ten years ? for the tablet market to even tend towards making this a duopoly. I mean, it is a monopoly now, the only difference is, its almost a monopoly (by consumer choice), sort of a natural monopoly - like our electricity transmission lines. Tells you how 'crucial' it is - to get there FIRST. Zuckerberg will tell us that.

5. Its no surprise that Froyo is going down - its a mock-OS. Fascinating article - make it a little shorter next time, just an honest feedback. I look forward to your thoughts, and kindly tag me - when you write more next time.

have a lovely week

LJS

@LJS: thank you for your kind comments! There's actually a ton of analysis that could be done in regards video consumption, and we're only starting to scratch the surface on this one using Big Data tools. :) But we will try to shorten the length of the articles next time; we wanted to make sure to present as thorough an analysis as possible the first time.

Some more comments on your points:
1. Power consumption: interesting correlation, thanks for pointing that out!
2. Kindle Fire: I suspect that there was the "new, cool device" phenomenon when people first bought the Kindle Fire and started using it for consuming video (among other things), but it may have died out - and as you point out, those users are likely just using the Kindle Fire for reading books or other light uses.
3. Connected devices: that might explain the heavy PS usage in the late night / wee hours of the morning!
4. Apple dominance: Indeed. I doubt that Apple will even close to vanish from the market :) and it tells you about the state of the tablet market. I think it will be interesting to keep an eye out on the upcoming Android and Windows 8 tablets and see if they radically alter the marketplace, or they just continue holding the candle to the iPad.
5. Android versions: Yes, and our analysis matches closely the statistics that Google is publishing about Android versions on their Play Store. Froyo is indeed really outdated at this point, and its a good thing that Gingerbread and its successors are gaining traction in the marketplace fast: it helps push the envelope on innovation.

Thanks again for your comments! We'll try to have more interesting analysis in future posts, but also try to keep it more concise.

absolutely Suman - thank you for your responses, completely appreciate your time - and the next time you write, I'll be back to discuss again. When you have a few minutes, let me know your initial thoughts on Jelly Bean ? and how good you think the upcoming in-house built version of google maps might be in iP5 ? I did play with Bean, but I'll tell you what I think after you've expressed your thoughts.

@LJS: thanks again! As for Jelly Bean: I now have a Galaxy Nexus and hence Jelly Bean - I think the UI (and video viewing experience) are a giant step up from previous Android versions (I upgraded from a Android 2.2 handset) but I do think it still falls a bit short of the Holy Grail of the simplicity and intuitive nature of the "iOS experience". Having said that, it is definitely a much better experience than earlier flavors of the OS.

As for "upcoming in-house built version of google maps might be in iP5": sorry, didn't quite catch that. Are you asking about the just-updated Google Maps app for Android (I got the update yesterday, haven't played with it yet) or Apple's in-house maps app for the iOS?

I referred to the in-built (Cupertino) substitute program for google maps - in iPhone 5. Its Apple's effort to de-google itself. Me as well - I really liked ginger bread, but I love Jelly Bean. I got my mum to play with both iPhone 4s (mine) and a Bean at the store. She is 55 years old and has always despised smart phones because she felt they were so disorganized and make her feel stupid. I just gave her a 4s, and she did not put it down for hours - she felt she finally dont need to ask me anything, and felt liberated. "Intuition draped in Elegance". She feels sad at times, that Steven passed away before she got to know about him. She felt Bean is fine - but there are still a couple of additional layers of complexity that she dont need. I rather find her feedback fascinating, because I know the iOS really made her feel - self sufficient. I am so interested in how Jelly Bean transmorphs itself into the future.

cheers - LJS

I think a discussion on which browsers people use - to do different online activities is also fascinating. We know Chrome outpaced IE - but Safari is still the most used browser in tablets. Hence just an idea, and it will be very interesting to see how that shapes up as well. The new safari is on par with Chrome, in terms of pace, but I do find the fact that you cannot use a picture online as your desktop background - in Chrome frustrating (the right click has no such option on Chrome). I dont know why I need to get a fix for that. I love the new Safari, but elegant in operation is as well. Love the elegance in its design, instead of the bulk robotics of IE, and to a lower extent - Fox.

Cheers -

Interesting stats.

It might take a lot more work to get, but I would be very interested in knowing about video encodings, especially bit rates. What are the most successful? Which are not? Where are the lines between those that always work and those that cause players to stutter and choke?

@Bob, thanks for your comments. We are indeed looking into analyzing more in-depth data, and some of the analysis we want to do does relate to video encoding, quality and user experience. As you can imagine, that will involve more data, more accuracy, and accurate analysis of the data we collect. We are definitely looking at how to best do this, and will definitely publish any interesting findings in a future blog post if we do.

I've just struggled through the wealth of stats provided from this a.m.'s September Email Newsletter.
THANK YOU VERY MUCH!
As a non-tech, non-little-device devotee, (don't even own an Ipad because I can hardly see the device itself), my interest is here:
REAL MARKETING
1. What age groups seriously use these Tiny PC Devices for anything except FaceBook/Twit/texting?
2. Has there been any legitimate study to show people of these groups buying *anything* except music via TPD's?
1. 5-15 years
2. 16-25
3. 26-40
4. 41-80

A colleague of mine in the real world of product marketing invested extremely heavily in the TPD market over a year ago. So it was really working in the dark.
EVERY figure he saw in marketing to TPD owners turned out to be hyperbole.

He almost went broke until he saw the difference between his FBS (Fixed Bigger Screen)split-market analysis and realised that almost nobody researched/used/bought ANYTHING via TPDs except Itunes.

He will be enlightened/saddened greatly by the data in just this month's Newsletter!

I remain, confusedly yours,
Himagain

@Himagain, we currently only measure video consumption on our sites/properties by people using smartphones/tablets. We don't have any data on whether the users of these devices are actually purchasing anything. As for other statistics, we are looking at other ways of evaluating the data that we have, and will certainly publish any interesting findings in a future blog post.

Post new comment

  • Allowed HTML tags: <code> <blockquote> <em> <strong> <strike> <ul> <li> <ol>
  • You may post code using <code>...</code> .
  • Lines and paragraphs break automatically.
  • Web page addresses and e-mail addresses turn into links automatically.

More information about formatting options