Android 2.2 Runs 450% Faster than 2.1

The new Google Android version 2.2 appears to be as much as 450 percent faster than version 2.1, according to test run by Android Police. To the extent that faster processing means lower latency for any number of operations conducted on a mobile device, that should be a good thing.

Android Police tests suggest, for example, that the HTC Hero gets a test score of about 2 million floating-point operations per second, often called simply "FLOPs."

The Nexus One running Android 2.1 gets about 6.5 MFLOPS to 7 MFLOPS.

The Android version 2.2 operating system seems to run at 37.5 MFLOPS. One practical result is that Flash-authored video should run much better, as Flash puts strain on processors. Much-faster processors should mean much-better video performance.

While real-life applications will most certainly not be 450 percent faster across the board, but it stands to reason that it will help most applications run faster.

link

Broadcasters Are Serving Up Lots of Web Video

Aside from YouTube, online video offered by broadcast TV networks and Web-only media brands, followed by magazine sites and music labels, seem to be getting the most traffic, a new study by Brightcove and Tubemogul suggests.

About 51.75 percent of viewers are navigating to video directly from the publisher’s main site. Google search drives 39 percent of viewing,  followed by Yahoo at 5.58 percent, Bing at two  percent and Facebook at 0.40 percent.

In the first quarter of 2010, the broadcast TV networks sampled in the study streamed 380 million videos, with Web media brands coming up close behind at 326 million video streams. However, the native Web brands, which include both video-only and general entertainment and news sites, saw 300 percent annual growth of video views in the first quarter, compared to 44 percent growth for the broadcast sites.

For all of 2009, Web media sites grew twice as fast as broadcast TV sites (165 percent compared to 74 percent). At this rate, they will overtake the broadcast sites in video views later this year, the study suggests.

In the first quarter of 2010, magazine-affiliated sites streamed 190 million videos, up 90 percent. In fact, magazine sites are streaming as many videos as music label sites, which came in at 191 million videos, up 60 percent.

Newspaper sites aren’t doing nearly so well, streaming 136 million videos in the quarter and growing five percent.  Newspaper sites are trying to catch up, though, and had two billion video player pageloads in the quarter (pages which loaded with a video player, but were not necessarily clicked on), compared to 1.2 billion for magazine sites, 760 million for Web-only media, and 670 million for broadcast TV sites.

But newspaper sites are having a real problem getting their audiences to watch videos, the study suggests.

For every two billion videos they throw in front of users, only 136 million get viewed (6.8 percent). Broadcast TV sites are getting 380 million views for every 670 million attempts (56.7 percent).
Even magazine sites are seeing a 12.7 percent hit rate.

But newspaper videos get viewed "to the end" more frequently than videos on other sites. The completion rates for videos on newspaper sites are 41 percent, versus 39 percent for magazine sites, 38 percent for broadcast sites, and 29 percent for music label sites.

Google Android Strategy is Working, Despite Nexus One

Google's strategy of seeding the market for its Android operating system, unlike its experiment with device retailing (NexusOne) seems to be succeeding.

The Android operating system "continued to shake up the U.S. mobile phone market in the first quarter of 2010," moving past Apple to take the number-two position among smartphone operating systems, according to The NPD Group.

Based on unit sales to consumers last quarter, the Android operating system moved into second position at 28 percent behind RIM’s operating system (36 percent) and ahead of Apple’s OS (21 percent).

Google's effort to retail unlocked, full-price handsets using a Web-site-only approach does not seem to be working out so well, as one might have predicted. Both Sprint and Verizon Wireless have declined to sell the Nexus One, though the logical explanation is that the HTC "Evo" at Sprint and "Incredible" at Verizon Wireless are functional equivalents, at the very least.

And it might just be the case that the battle between AT&T and Verizon Wireless accounts for the change, as iPhone sales in the United States are exclusive to AT&T, essentially limiting sales, while Android devices are pushed both by Verizon, T-Mobile USA and Sprint.

Verizon also has been aggressive about offering "two for the price of one" sales of Android devices.

Smartphone sales at AT&T comprised nearly a third of the entire smartphone market (32 percent), followed by Verizon Wireless (30 percent), T-Mobile (17 percent) and Sprint (15 percent).

Exclusivity on AT&T’s network obviously limits the potential sales for Apple to some extent. Verizon has more than 92.8 million subscribers, none of which can buy an iPhone for use on the network.

It isn't so clear whether the range of Android models or prices are a meaningful contributor to Android sales volume, but one has to think so.

The NPD Group cites an average smartphone price of $151 in the first quarter of 2010, roughly half of the $299 price tag for a top-shelf iPhone. Apple offers subsidized models at $99 and $199, but most subsidized Android phone prices top out at $199 and go down from there.

The Samsung Behold 2 running Android is currently free with a service plan at T-Mobile, for example. With so many choices, consumers can find Android units for well under $99 these days and can shop around in a greater range of price points.

link

Verizon LTE Test Runs at 8.5 Mbps

In recent tests, Verizon Wireless has found that its new Long Term Evolution network runs at about 8.5 Mbps in the downstream direction in the real world.

Web Apps Will Catch Mobile Apps, Study Suggests

Count Global Intelligence Alliance as among the analysts who believe Web applications will be a viable competitor to app store programs over time, and that content distribution is likely to be a direct beneficiary of the trend towards using Web browsers to serve up mobile apps.

The same might be said for subscription-based mobile services such as news and weather as well.

Despite conventional wisdom, by 2013 HTML5 will enable Web-based apps to provide user experinces that rival that of mobile apps, GIA argues. But there are other reasons to believe Web-based apps will prove attractive. Web apps offer an architectural advantage, namely cross-device launches.  Mobile apps have to be adapted for each operating system, and often for discrete devices as well.

The Web also arguably is a better platform for subscription-basedservices such as communications, news, weather, financial services, retail and shopping, where user analytics are important. But GIA notes that smaller providers and pay-per-download services might well find the mobile apps route profitable, as such an approach can be directly tied to a clear revenue model.

But Web-based mobile apps will take a couple of years to develop. Right now, respondents surveyed by GIA say user adoption is about twice as high when using a mobile app approach. Some 47 percent of respondents reported that user adoption was higher when using a mobile app approach, compared to about 23 percent of respondents who said a Web approach produced higher end user adoption.

Web apps, on the other hand, are a bit more "sticky" than native apps, respondents report. About 27 percent of survey respondents said user activity peaked at initial download and then steadily declined. Only 15 percent of Web app developers said that was the case.

Likewise, about 23 percent of respondents indicated that user activity kept growing after download, compared to about 33 percent of Web apps users. Of course, that might be a statistical artifact produced by the different use cases.

To the extent that a mobile app provides access to "static" content, usage would decline over time, in much the same way that any user's viewing of a new movie will be highest at download and then drop off. Compare that to a cable TV subscription or news feed, that might be used on a continuing basis because the actual content is dynamic, rather than static.

The survey also found that end user session lengths tended to be longer for native apps, compared to Web apps. About half the respondents say native apps produce longer sessions. Only 20 percent of developers say Web apps produce long sessions. Of course, much depends on the type of application.

Many interactive or transactional apps will tend to last longer than many content delivery apps, if only because the transactional app will require time-consuming search and research. A user investigating air travel or lodging in a distant city might need to spend quite a bit of time conducting research, compared to a user playing a game or downloading a specific bit of content.

About 53 percent of native app developers reported that this approach cost more than creating a Web app, compared to 17 percent of respondents who said the Web app cost more than a native app to create.

About 43 percent of developers reported that maintenance and update of native apps cost more than a Web app approach.  About 24 percent of respondents indicated that a Web app approach was more costly to maintain and update.

About 60 percent of developers reported that a Web app approach was faster than a native app development approach.

Apple's iAd Shooting for the High End in More Ways than One

The conventional wisdom likely is that the new Apple iAd network is going to be positive for mobile advertising, showcasing what can be done with rich media on an easy-to-use device with a large screen.

But iAd might have another effect: driving up ad pricing double to triple current rates. That might be welcome for ad sellers, but not for buyers.

Since Apple never likes anything but the "premium" position in any market where it competes, that might not be too surprising.

According contentSutra, iAd cost per thousand impressions will be highly variable in terms of cost, but in some cases could wind up being triple what marketers are used to paying for banners, and double the price of a video ad on mobile (click on the image for a larger view).

In other cases, a very successful campaign that generates an unusual number of clicks could wind up being more than seven times what an advertiser had anticipated.

Keep in mind that Apple is setting the minimum annual spend at $1 million for brands to use its iAd platform. To put that in perspective, consider that Jaguar and Land Rover in 2009 spent about  $1.6 million in aggregate for mobile marketing.

Some advertisers might decide they like the iAd formats, and start shifting lots more money into mobile advertising. Others will push back against the price and take a wait-and-see attitude. Either way, iAd is going to have complex impact. It likely will spur larger mobile advertising commitments from some buyers more concerned with cutting edge than return on investment. Most will watch and see what happens, at least in terms of the high-end, showcase content Apple wants to show it can produce.

Clearwire Says It Has 3 Customer Segments

Clearwire says three customer segments already have emerged for the firm's fourth-generation wireless network. "A Clear customer is often a cord-cutter," says Clearwire CEO Bill Morrow. "This customer has a mobile phone, but no fixed line phone in the house." That customer fits the profile of an on-the-go user that wants broadband available outside the house.

"They also probably don't have cable TV service," he says, and tend to substitute Hulu, YouTube or other video sites for TV. It probably goes without saying that this sort of customer is single and younger, without children.

The "cable customer," on the other hand, "can't cut his Internet connection at home," says Morrow. That is because this second sort of customer has a family sharing a single at-home video entertainment connection, as well as a broadband Internet access connection.

In this segment, the 4G wireless service is part of a triple play package.

The third segment consists of mobile customers served by Sprint that wants to add a broadband access service for PCs and other devices, and for whom the convenience of having both mobile phone and broadband access services on the same bill, provided by one carrier, is valuable.

"When you go into the market you see all these different customer segments and it's easy to see that you can get more subscribers in total by marketing to different groups instead of just having one brand," says Morrow.

Clear customers do use much more bandwidth than the typical 3G user, though. "We're finding that customers are using on average 7 GBytes of data per month on our service," says Morrow. "The average amount of data a 3G subscriber uses a month is about 1 GByte to 2.5 GBytes a month.

It is not certain why this is the case. It could be that heavy users are attracted to the unlimited access, with no monthly caps. It could be that knowing there is no cap motivates usage, as broadband leads users to consumer more data than they do when on a dial-up service.