Photo: Yummy Scummy by Zach Manchester – Some Rights Reserved

There’s been quite a stir lately about the viability of the Web on mobile devices with various folks pronouncing it inferior especially in comparison with mobile applications.

Yesterday Malcolm Murphy at Mobile Industry Review blasted; “Can we all admit that ‘Mobile Web’ is total rubbish? ” He pointed out that there’s a huge gap in performance and usability between web apps in desktop browsers and those on phone.  Specifically;

• Mobile browsers are frustratingly slow and don’t do Flash or JavaScript well, if at all.
• Sites designed for mobile might give a better experience but are impossible to find.  Do you go to  m.site.com, site.mobi or site.com/mobile ?

Malcolm claims mobile applications are the way to consume web services on phones.  He says that he and the people he knows do 90% of their interaction with web based services using mobile applications like the Gmail app rather than a mobile browser.

While Malcolm grants that the iPhone delivers a superior experience, he dismisses it as a “niche” device and says that the popularity of  Facebook and Twitter apps even on the iPhone further validates applications as the preferred delivery method.  He admits that mobile browsers will get better but will always be inferior to desktop ones with the implication that the mobile browser will never be worth bothering with.

Malcolm’s viewpoint was echoed in reports from the MobileBeat conference last week, where  GeJar AppStore  CEO Ilja Laurs was widely quoted as predicting  that “”Apps will be as big if not bigger than the Internet”

I actually agree with  Malcolm on many of these points. Mobile browsers are inferior to desktop browsers and probably always will be.  Well designed mobile applications do generally perform better and give a superior user experience compared with their mobile browser based equivalents.  However I strongly disagree with his (and Ilja’s) conclusion that the majority of web use on phones will be with single purpose applications  rather than web browsers.

Applications are important and will be increasingly popular with users. There are some types of services like navigation and mapping that will probably always be better as done an app.   But it is the browser  where the real growth in the consumption of cloud based data will occur.  The reason is scalability. Apps are not scalable on a couple of levels.

First there is the scale of the web itself. There are millions of sites.  Sure the average person visits only a few dozen sites with any regularity.  But the web is built on hyperlinks.  We use it by following links to other sites.  I’d be willing to bet that by following links most web users visit at least a couple dozen new sites each week in addition to the two dozen they follow regularly.  Users who are active on forums, Twitter, Facebook and other social networking sites where link sharing is popular may visit a hundred different sites every week, thousands a year.  Do all these sites have apps? Of course not and even if they did no one wants to install a separate app for every new site they visit.

The other scalability issue is the one faced by app builders.  Mobile is very fragmented.  Building and maintaining apps for every mobile smartphone OS and every subtly incompatible variant of Java ME is incredibly expensive.  No want can afford to do that.  Google VP of engineering Vic Gundotra agrees saying at MobileBeat that even Google is “not rich enough” to build apps for all platforms. Gundotra believes that the future is  browser based mobile apps. Even GetJar’s Laurs seemed to agree with Gundotra on app development costs saying;  “It is fashionable to do apps and every media outlet tells you apps are cool. …But the economics are a different story. The ratio of those developers who will fail is about 90%…”

So getting back to Malcom’s point that the web on mobiles is rubbish.  Yes there are problems, the user experience is generally not very good. But everything that’s wrong with the web on mobiles is fixable.  And it’s being fixed right now. There is a browser war going on between Apple, Google, Nokia and Opera that is driving the performance and usability of mobile browsers ever higher.  Apple started the war by defining a new level of mobile browsing speed and  usability  with the original iPhone and improving on it for the 3G.  Opera has been building great mobile browsers since 2000 and with Opera Mini has brought an  iPhone-class browsing experience to almost every device including inexpensive “dumb” phones.  Nokia is very much in the game too. Rafe Blanford at All About Symbian just released the results of a browser speed test where he found the latest Nokia WebKit browser on the N86, N97 and 5800 was significantly faster than the iPhone 3G.  Mathew Miller at ZDNet’s Smartphones and Cell Phones added data to Rafe’s showing that the iPhone 3G S’ browser tops Nokia’s latest in speed but is in turn beaten by the latest version of Opera Mobile. Miller also called Opera Mobile’s usability better than that the iPhone browser.  And Rafe’s test found that server assisted browsers like Skyfire and especially Opera Mini are the fastest of all.

As for Javascript, it’s already supported by Opera Mini and Mobile and by the Nokia, Android and the iPhone browsers. None of these offers desktop level JavaScript speed and support but they are good enough that  many AJAX desktop sites work well with these browsers.  Google’s latest iPhone and Android web apps make extensive use of JavaScript and run in Opera and the Nokia browser too. As for Flash, I see a lot of performance and usability issues with it even on the desktop. HTML5 will soon provide better, more portable ways to do video and animation than Flash on the desktop and mobile. Until then, Nokia’s browser already supports Flash video and full Flash 10 support is coming to Android, Palm Pre and Symbian latter this year.

As Malcom points out, most desktop sites are hard to use in small screen browsers and discovering the mobile equivalents of  many popular web sites is difficult.  The answer to this is greater awareness of mobile by web designers, developers and publishers.  Every study of mobile web usage  I seen shows it rising rapidly around the world, particularly the developing world.  According to Tomi Ahonen, there are now 1.05 billion users of mobile browsers worldwide, which is slightly more than the number of desktop web users. Mobile is becoming the tail wagging the Web dog. Web publishers and developers who ignore mobile do so at their own peril.  Today, most mainstream websites are doing some form of adaptation for mobile browsers. Much of it is rubbish but it’s bound to get better with the increased awareness of mobile that is developing in the mainstream web design and development community.

The discovery problem is fairly easy to fix.  The answer is browser detection and adaptation with thematic consistency.  What this means is that if you visit a given URL with different devices you should see more or less the same content but formatted for an optimal experience on your device.  Mobile and desktop versions of the same content on the same link, not a different one.  Visitors to  site.com or site.com/some-really-long-deep-link/ should  see essentially the same information in  mobile devices that they do in a desktop browser.  Adaption will apply formatting changes (single column, reduced image size, pagination if needed) to make that content usable on the mobile. That is thematic consistency and it’s what will make mobile discovery transparent and thus easy for users.

There is one danger with browser detection and adaptation that must be avoided. Browser detection is inexact and it’s dangerous to make assumptions about whether the user wants mobile formatted or desktop content. Sites using browser detection to automatically deliver mobile formatted content to browsers should always include a link pointing to  the full-web version of that content if one exists. Once the mobile visitor switches to the full-version that switch needs to be persistent throughout the current session. The user asked for non-mobile formatted content, clicking a link on the full site should not dump them back on the mobile site and vice versa.

In summary, Malcolm is right, the web on mobiles is frequently rubbish.  But the problems are fixable and they are being worked.  The alternative of accessing web content using applications is not scalable.  For a services and sites that one visits frequently, particularly highly interactive ones like email,  Facebook or Twitter, an app will often provide a better user experience on mobile than a web app. But that’s not always the case.  For me the  iPhone/Android/Symbian web version of Google Reader beats any stand alone mobile RSS reader application.  It’s extremely fast, easy to use and keeps my reading history and status synchronized with Google Reader on the desktop. Dabr, a mobile Twitter web app, rivals the best dedicated Twitter apps on most platforms.  Most of the web content I consume on my phones isn’t available in an app format and probably never will be. The web is huge and diverse.  There can never be enough single purpose apps to cover it all and if there were they wouldn’t fit on my phone. The real growth in mobile data adoption will be web browser based.