Blog Post

Using the Mobile Web is a Sticky Proposition

Stay on Top of Enterprise Technology Trends

Get updates impacting your industry from our GigaOm Research Community
Join the Community!

Another Mobile World Congress, and another week full of promises coming out of the wireless industry that we will find eventually use our mobile phones to access the web much like we use our PCs. While I do believe we’re closer (and some give the iPhone credit for this), I still think the finish line is far off. The cage match du jour, the fight between Linux operating systems offered by Google and the LiMo Foundation, underscores one of the big difficulties of using the mobile phone for a rich Internet experience.

There’s too much variation in operating systems and end devices, which makes it hard for developers to build applications for a mobile phone. Obviously people recognized this when it came to building applications for social networks (see: Open Social), but efforts to build platforms that span mobile phones are nascent.

One problem lies with the hardware, which varies from simple phones that serve mostly to make voice calls to smartphones that have the processing power to handle Office documents or broadcast television. There’s also an interface issue, such as whether it’s a touch screen, a scroll wheel, stylus or keypad.

Seriously, aside from finger cramps, anyone using a keypad to navigate the web is going to get really frustrated really quickly. Designing a browsing experience and services to optimize so much variety general results in designing for the lowest common denominator, or cutting them out entirely, and sticking with the few that have smartphones.

But the big problem is software — there’s too many operating systems to choose from. So the open platform zeitgeist is going mobile. AOL unveiled its open mobile developers platform earlier this week at the World Congress, based on technology assets it acquired from Airmedia, so it won’t be real until this summer. The platform allows a programmer to build applications for up to 150 different handsets using a variety of operating systems, but requires a client on each mobile phone.

Also trying to make the development side easier is Streamezzo, a Parisian startup that has raised $48 million to create its cross-platform software development kit. The kit won’t be available until Feb. 25, and requires a client on the end user’s handset. SFR, France’s second-largest mobile carrier, is running applications built on the Streamezzo SDK. Yahoo has launched a mobile widget development platform as well that went live this week, after being announced at CES in January..

On of the more interesting approaches is being taken by Chicago-based Novarra, an eight-year-old company that is working with carriers including Vodafone, U.S. Cellular and 3 Hong Kong to deliver the web to any phone, even low-end handsets. Novarra offers an appliance for carriers or a service that essentially offloads 80 percent of the data processing associated with downloading a web site to servers run by the carrier or Novarra. This cuts down on the amount of data traveling over the carrier network, and makes load times faster. Content providers such as Yahoo also use it to deliver lighter applications for mobile phones. Novarra powers Yahoo’s oneSearch via mobile.

Novarra’s success at driving data usage among its customers’ end users is exhibited by an increase of between $5 and $15 in ARPU for the carriers deploying the Novarra software. One only needs to look at AT&T’s recent profits, which were driven by wireless growth, to realize that pushing easier access to the Internet for all will drive revenue for carriers. The key is making it as convenient to use the web on a mobile phone as it is to use it from a computer.

21 Responses to “Using the Mobile Web is a Sticky Proposition”

  1. BTW – Opera Mini implements a lot of features that I wish my desktop browser had… Seamless backtracking without reloading, etc.

    The Opera desktop browser already does that. Since your visited pages are stored in the browser’s cache, Opera doesn’t need to download the page again.

    Opera Software

  2. I allready have web 1.0 and part of web 2.0.
    (excluding pages heavily reliant on flash and some of the more heavy web 2.0 javascript apps).
    I use Opera Mini on a Nokia E61i with 3G network.
    A large proportion of my personal surfing happens from my smartphone at whichever location is convenient.

    Why do I want a mobile web?
    Did anyone want a unix-web?
    A windows-web?

    The core design principle of the web (departed from in too many cases) is complete platform independence.

    A nod in the way of easing rendering for a small device is sometimes nice (more often just irritating as the device renders the original page better that the page adjusted for mobile in an “attempt” to help it).

    The problems you describe are problems for the application developers (mostly browser designers) and are not an insurmountable impediment.

    BTW – Opera Mini implements a lot of features that I wish my desktop browser had – automatically focusing on the main text-body part of the page. Full Page overviews. Nice Transition animations. Seamless backtracking without reloading, etc.

    (I live in Africa)

  3. Few brief points;

    1.) Novarra are the worst thing to happen to the mobile web in it’s short but promising history – seriously. There is an ongoing battle in Europe and elsewhere between developers and carriers that deploy Novarra technology. Not only do Novarra arguably do a REALLY bad job of transcoding, but the key point is, they block standard long-accepted header information sent with each web page request, that identifies the model of phone connecting. Blocking this means developers can’t deliver the right content to the right phones, which results in broken sites and services, with much revenue loss. Novarra and the carriers have deliberately avoided this issue and lied repeatedly about addressing it.

    2.) This article is not so much wrong, but highly US-centric, and thus heavily blinkered. Before the iPhone was a glint in Steve Job’s eye, there were more people accessing the web on mobiles elsewhere in the world, than the number of iPhones that have been sold or will be for a long time. The iPhone has simply thrown the mobile web into clearer focus for a nation (the US) that has had terrible mobile services compared to Europe and Asia for too long. In the US, you don’t really know the great experience of the latest Nokia web browsers for example, built into S60 powered handsets in tens of millions of consumers hands. Aside from touch, we’ve been having great web experiences built on the same Webkit platform as iPhone Safari for far longer than iPhone users.

    3.) Linux vs. Limo is a minor skirmish in mobile at the moment.

    4.) The plethora of different runtime platforms is not that big a deal. Increasingly developers will focus on common delivery platforms on top of the OS – namely, widgets and web, possibly J2ME. Widgets, if they are standardised as looks likely, are a great hope for a common platform for the industry.

    Alex Kerr
    London, UK

  4. For some reason you’re missing the second half of the crux of the matter. Sure, UI is an issue, but if mobile users were able to download content quickly and cheaply, it would be less important. However, existing mobile data is not priced transparently for the average consumer (how much does downloading 3 web pages cost?) and even in Europe, is very slow. In the US it must be absolutely glacial. These 2 factors mean that the user experience and the quality of the web pages become the final straw to mobile data use.
    The iPhone’s real shift is that it is sold on a data plan, and so users feel that they might as well use it for the internet. Even 5 years ago, my Sony Z5 had a decent html browser onboard, so how webpages look is really not the most vital factor in mass uptake of mobile internet.

  5. To most web developers and those of us who care about our audiences enough to design mobile web sites specifically to cater to mobile usage, Stacey’s characterization of Novarra as a success is unfortunately very one-sided and very incorrect. While it perhaps did drive data usage for mobile phone companies, the damage it has done to the existing mobile web is not to be neglected. There is no mention of that in the article, and I find that very disturbing. As David above mentioned, I have documented extensively how Novarra is destroying existing sites that are designed for mobile.

    There are numerous other examples that Novarra is trampling all over the mobile web without regard of existing standards and practices.TeliaSonera in Sweden is using Novarra to transcode sites found through their search. The questionable practice is not only they are transcoding third party sites without the site owner’s consent, they are running ads on those content. I am not sure if there are rev share agreements with the third party sites. But if they are running ads without the content owners consent or a rev share agreement, it would be a terrible mistake.

    If you need a good illustration that there is no escaping of designing for specific mobile use instead of just throwing a transcoder in front of the desktop web, see Barbara Ballard’s excellent presentation about leveraging device characterisics:

    Miss Higginbotham, I hope you can write another post and give equal voice to the other side of the debate — the fact that Novarra and their wireless carrier customers are destroying the usability of the mobile web, and in some case doing something questionable in deploying their technology.

  6. Many who didn’t really “get” the desktop web thought it was just a new way of displaying printed material. Hence brochure-ware, to the amusement and despair of the web in-crowd.

    It’s happening again. The mobile web is far more than just the sedentary web viewed through a keyhole. It is different medium which brings new opportunities and requires new philosophies.

    James Pearce
    CTO, dotMobi

  7. If we replace “mobile” with “desktop”, we’re back to the late 70s and early 80s, before the DOS – and later Windows – hegemony drove the industry toward broadly accepted standards.

    Exciting times, especially given the number of open standards currently being pushed as lingua franca solutions to the current conflicting/non-existent-standards dilemma afflicting the mobile world.

    I feel the need to pull up a seat, grab hold of my mug of tea and prepare for what promises to be an entertaining, longish show. Neat.

  8. Jesse Kopelman

    The big problem is not really software, especially OS. After all, browser as OS is a perfectly good solution for a device that will always be underpowered compared to a desktop (thanks to battery dictated power constraints). The big problem is user interface hardware. We’re a long way from some sort of usable minimum resolution on the handheld being as ubiquitous as 1024X768 is on the desktop. Even after everyone is at 320X240 or (hopefully) better, what about input method. Not everyones got a touchscreen. Not everyones got a keyboard. Meanwhile, are any of these things part of the proper paradigm for the mobile web? Realistically, we are 10 years behind the desktop, when it comes to the mobile web. There is enough potential for profit, that I’ve no doubt the breakthroughs will come. The only question is when. Linux on the mobile is a great thing, as it gives the best playing field for wisdom of the nets to be brought to bear on these problems.

  9. kurukshetra-wallah

    The assumptions that mobile-web = PC-web or mobile data device = PC are flawed and excessively US-centric. The analysis, in turn, suffers from these flaws though consistent with its own logic.

  10. Great post Article. It’s good to see people are finally starting to pay attention to Mobile Web content.

    For too long it’s been an ‘also ran’ repurposing of desktop content that content providers threw up at the last minute without even bothering to monitor traffic (didn’t matter most of the time as the content was static and unchanging anyway).

    I’ve said for a long time, if someone visits your mobile site and you dont have any analytics and dont know anything about them or their visit…..does it count (as a homage to the saying “if a tree falls in the forest and no one is around to hear”).

    If you run a mobile web site check out Amethon’s Mobile Analytics from

    One of the worlds first analytics applications specifically built for mobile browsers.
    With no page tagging, artifacts or javascript we offer a real time analytics solution with no overhead or lag.

    You can finally see what your mobile web visitors are trying to tell you about your mobile content…….good or otherwise. At least then you will be able to move forward offering dynamic mobile web content that interests your visitors – or is ‘thumb’ fatigue a bigger issue than WAP content providers think.

    Dean Collins

  11. Why surf the web? Why can’t we just talk to the phone to do this for us. I believe we are still stuck in trying to cram the whole desktop into the phone. I don’t want to surf the web on my mobile, I want to experience the web which is quite different.

  12. Stacey,

    I think that you’ve made an excellent observation, one I think reflects the emerging 1st generation of mobile web adoption in North America. Particularly, your assessment likely rings true for thin client applications as opposed to web based services. For the wired internet, many early products were pc based clients – think Prodigy, CompuServe, AOL, and many others. Today the web is strongly represented by the web services model, of which open platforms embody. This said, many of the issues that you correctly highlight will mitigate as the web service model, or “mobile web services” (MWS)model is adopted and deployed.

    Note that the iPhone is the ultimate embodiment of the MWS model, albeit in a closed development environment. This model is a much better model for consumers in that it does not require downloading and knowledge of the cell phone operating system.

  13. I’m sure the same issues were raised by those who were planning the Internet before its inception. The problems with multiple browsers, operating systems are an analogy to touch screen etc. and mobile softwares. Services which try to solve the problem by having people install their own clients are approaching the problem the best way they can.

    The solution lies in the standardisation of the mobile web, along the same lines of existing standards for the Internet. As you pointed out, it was with the iPhone that people ‘actually’ began to use a mobile phone to surf the web, so it’s evident how new it might be to people. Developers never have any incentive unless they see people flocking towards something. Now that they can, I’m sure companies will make efforts to come out with mobile versions of their applications/services.

    What the mobile operating system makers should do is improve the included browser software (and not rely on third-rate Java based browsers), so that they have a wider tool-set to be able to offer the maximum they can. Safari (on the iPhone) and Opera are good choices at this point. I see majority of the effort lying not with web-application developers, but phone browser programmers.