Dec 18 AT 3:47 PM Brendan Nelson 24 Comments

Are mobile apps here to stay?

A few weeks ago a guest speaker came to our office to talk about mobile apps. His company produced a lot of them, for pretty big brands. He knew his stuff: the team here was both impressed and engaged.

But an exchange during the following Q&A session stuck in my mind later. One of our directors asked a question: is the mobile app destined to be a transitory phenomenon, something that will fade away as mobile browsers become capable of delivering the same functionality?

The speaker was adamant that this was not the case and that mobile apps were here to stay. He felt that Google’s increasing preference for mobile browser apps over native apps was misguided and that Google were wrong on this one. Mobile browsers were so far from rivaling the functionality of native apps that it wasn’t even worth thinking about.

I was tempted to counter this point by bringing up the iPhone’s support for HTML 5 and starting a detailed discussion about in-browser capabilities. But this wasn’t the main subject of the talk and I’m in no way an expert on HTML 5, so I decided to keep my mouth shut instead.

In the weeks since the talk, however, I’ve often found myself turning this question over and over again in my head. And the more I think about it, the more I feel that mobile apps are basically doomed — or at least I hope they are.

Don’t get me wrong — they play an important role. It’s good that so many people today see phones as devices for more than just calling or texting, and the iPhone and its suite of native apps is largely to thank for this. But in the longer run, the publication and distribution model they are based on has to go.

The idea of tying software to a single hardware platform is anachronistic, uncompetitive and limits user choice. This is bad enough when you’re dealing with computers, but it’s even worse when the devices are as personal as mobile phones. People should be free to choose a different phone without needing to buy new versions of the software tools that have become integral to their lives.

Aside from user choice, there’s a more practical reason why the native app model is unsustainable. Developers won’t want to keep maintaining multiple codebases for the apps they produce, especially when there’s the option of building an equally functional in-browser app which any standards-based client can run. And although Apple might hope to render this point irrelevant by establishing monopolistic domination of the smartphone market, relieving developers of the need to consider other platforms, current research indicates that they won’t succeed.

The smartphone OS market will be more fragmented in 2012 than in 2009.

The smartphone OS market will be more fragmented in 2012 than in 2009.

A more fragmented smartphone OS market will increasingly compel developers to support separate codebases for Windows Mobile, RIM, Android, Symbian and the iPhone. But as mobile browsers become capable of delivering similar interactivity, serious developers will become inclined to start using the browser as the platform instead. This will be a good thing for users and the industry alike.

If I’m correct and native apps do fade away over time, we may look back on the era of pointless mobile apps as just one among many strange blips in the history of technology. But despite some early rumblings from notable developers, native mobile apps will be with us for some time yet — and, in the medium term at least, they still have an important role to play in encouraging mainstream adoption of the mobile internet.

As head of strategy at interactive agency Tobias & Tobias I get to work with lots of clever people. My current obsession is the potential of search and data services in the real-time web. For more from Brendon, visit his blog.

    Most Tweeted This Week

  • http://Website Gilbert Orbea

    Is it wrong that when I see the iPhone decline and Android grow I can’t help but smile?

  • http://Website Yatima

    I think you are partly wrong on this. The whole point about native apps it to gain access to hardware functionality as well as the ability to run the app without network connectivity.

    • Zacqary Adam Green

      You mean like HTML5 can?

  • http://Website lordhong

    This only applies for networked apps, especially SNS sites/apps.

    For hardware dependent apps, like 3D games, you still need to get as close to the hardware as possible.

    Also apps using sensors, such as the accelerometer.
    Look at what an EPIC FAIL job Palm did to webOS with accelerometer. You only get about 4 reading samples a second, comparing to about 100 in iPhone SDK. You can’t build any playable games like Labyrinth.

  • Some Random Chris

    I’d be concerned about speed. For now, I much prefer native apps, since the Android browser is super slow (non-Droid). Plus, a network hiccup can bomb out something you’re working on, while a native app would cache it for sending later, etc.

    Google may want apps to all be web based (greater penetration), but if that was their actual philosophy, I wouldn’t be reading my email in Android Gmail, listening to podcasts in Google Listen, or taking picture searches with Google Goggles.

    I think in the end it will be some native, and some web. It’ll all be app specific. There is no such thing as “all or nothing”.

    • Hugh Isaacs II

      Overtime you’ll see this change.

      The web now won’t be the web later, Google and other companies are continually adding features to browsers bringing them closer and closer to native applications.

      Google even has a project called Native Client allowing native applications in the web browser just for this purpose.

  • http://Website Josh

    OTOH, I already use Twitter’s mobile site instead of Twidroid, (I found it starting background services even though I’d told it not to in preferences.) and an app that’s just a wrapper for a website’s about 80% pointless. (Network connectivity causing problems is the reason it’s not 100%)

    Still, performance, native look ‘n feel, and the ability to do cool things without waiting for the W3C to settle on a standard are all good reasons to go native.

  • http://Website Quasar
  • Gregg

    Apps will not go away, for the reasons londhong states above. However, what you will see, are less people, particularly businesses, creating apps just to have an app. Let’s face it, when you want to search for something using your phone, what do you search with, Google or the App Store? People who get their traffic from Google will be better served with a web app and not a native app. Using Google to search has become a part of human behavior that isn’t going to change anytime soon.

  • http://Website Zer09

    Web applications have a hard time replacing native applications on desktops with more powerful processors, larger screens, and more natural input methods. Computing moved away from the mainframe-dumb terminal model, and now it is moving back, but its only a matter of time before the pendulum swings back the other way. No one way is the future it will always be a blend.

  • Hugh Isaacs II

    The thing is, what a lot of people don’t see is that the future of the web is to be able to do everything native applications can do.

    It’s like Google even said, for now native code is needed on smart phones but over time that’ll change.
    Everyone’s always saying “the web can’t do X” but yet several companies are working on getting it to do just that soon.

    There’s specs in the W3C for cameras, microphones, file system access, installing web apps to the desktop, plus Mozilla and the Khronos group are working on OpenGL for the browser and Google themselves are even working on native code support and notifications.

    And Apple already has animations and 3D layout support on the mobile Safari (View this in your iPhone browser:

  • Brendan

    @Yamita – that’s a good point, integration with specific hardware features is indeed a huge benefit of native apps. Mobile devices are still primarily used for communication but as that changes more sophisticated use cases will offer a niche for native apps. Someone editing audio on a mobile device is unlikely to be using a browser-based app, for example.

    @lordhong – you’re right, certain types of app are definitely set to remain in the OS domain for a long time. And while I can picture a future where closer browser-OS integration provides browser apps with standardised access to sensor APIs, that future is some way off. Adobe would like to move in that direction ( but to be honest I don’t have much confidence in mobile Flash.

    @Josh, @Some Random Chris – I’ve started using the new Twitter mobile site instead of Twidroid too, but agree that the browser in Android remains way too slow. If the speed problems are resolved though I’d be happy to use the browser versions of Gmail and Calendar – the features are good, it’s just the speed that lets them down.

    @Gregg – I agree, apps created just for the hell of it will not be around for much longer. Maybe this is me being too cynical but I sometimes see the native app bandwagon as the new Second Life; businesses see competitor activity in the area and feel compelled to jump in too, with not enough thought given to what their users might get out of it.

    @Zer09 “No one way is the future it will always be a blend” – no truer words have ever been spoken.

  • William Furr

    I think lots of people, including your guest speaker and many of the commenters here, are under-estimating how close web apps are coming to native code performance. They’re getting closer every day, to the point where they will be indistinguishable to users.

    HTML5 brings many of the key native performance features that desktop apps use to web apps, including real-time screen rendering with canvas and javascript.

    One limiting factor left is real-time access to sensors, as mentioned before, and that’s changing too.

    On the development side, targeting HTML5 browsers with a web-app has so many advantages over native app development that it’s crazy.

    Patching becomes ancient history and you have the ability rapidly iterate your software and test out the changes. Putting software in front of a user as early and often as possible is the best way to make good software, and Google knows this. Just check out the iterative web app series by the mobile Gmail team.

    You also automatically get cross-platform support by any HTML5-compliant browser, and *this* version of HTML is a much stronger spec than any earlier version. Many of the cross-browser compatibility issues are a thing of the past, and HTML5 even degrades nicely on non-HTML5 browsers.

    Particularly in the case of consumer information appliances, like smartphones, netbooks, laptops, even many desktops; web apps and a browser-centric OS are where they’re headed.

    For myself, there are exactly two things I do with my laptop that don’t (or couldn’t) happen in a browser right now: development with Eclipse, Netbeans, Flash CS4, and Visual Studio; and 3D games (I want to get in on the next WoW expansion).

    And in those two areas, there’s work on exposing OpenGL ES 2.0 through the Canvas element, allowing hardware-accelerated 3D graphics on the client and some interesting work on web-based collaborative IDEs.

  • Scott Kellum

    I always go back and forth on the idea of web apps. In the end my biggest concern is how will I listen to music or watch a video in the subway? I can’t comprehend HTML5 storing large amounts of data like this. As far as games and other proformance heavy apps go I have faith in some of the new technologies coming out like webGL. Browser testing on mobile phones is also a lot easier than on PCs due to the lack of IE and abundance of Webkit based browsers.

    Cool stuff to think about but I’m not converting to a web only os until media players and other apps work in dead zones.

  • Pingback: Haben Apps eine Zukunft?()

  • Pingback: Android and Me()

  • http://Website Zer09

    @all opponents of Native Applications

    I’m on an airplane and I want to play a game to entertain myself
    I’m on a hike in the wilderness and want to shoot some video of nesting birds
    I’m on a boat in the ocean and want to work on a novel I’m writing.

    Java, and Flash made claims to be able to have cross platform compatibility. HTML5 and Web Applications are just the next claim set to fail because Native Applications will always hold an amount of performance, security, and availability in some areas.

  • http://Website Tim

    As phone hardware and web standards continue to advance more _network based_ applications will become web based, that’s for sure, others never will, like the Compass app – even if you could make that web based (which would require someone to put in an API to access a devices compass into HTML!!!) why the hell would you? I don’t want to have apps I can’t use without web access and while you’re at it, those I do I don’t want to load the code for an application off the web, i’d prefer an app which only needs to load data from the site, not code as well. It’s not hard to see that loading the text of every tweet plus whatever encapsulation they have is going to be a lot faster than loading that and all the layout code for the webpage as well. Plus with HTC Peep I can catch up on tweets at work where I have no signal.

    Even if you could make a webapp with 3D rendering run native on a device it will end up running a lot better if it’s written specifically for that device manually since the web version built with a toolkit won’t be as optimized which is very important for something as resource hogging as 3D (or fancy 2D effects for that matter).

  • C. Enrique Ortiz

    This argument is not new, and is a pretty religious argument between web vs. local/native developers. Local mobile apps are here to stay, for a long time. Web runtimes will help the transition, once all access to handset capabilities are possible (and standardized), but I am believer that both classes will remain and co-exist. Where web apps suffices, web will do fine, but when wanting to build very rich apps, local apps is the way to go. Also, a better way to monetize mobile web apps is needed.


  • http://Website Bradley

    One tech you have all forgot is the coming of AIR 2.0 for mobile, with Flash Player 10.1 for mobile as well. This is really the one differentiating factor. When this arrives it will be available for all smarphone OS’s except iPhone, but i got a feeling in order to not lose market share Apple will have to bring it to the iPhone. This is the one place where fragmentation does not play a part with the possiblilty for local specific functionality like currently available in each native OS framework.

    Good time to be a Flash Dev :)

  • http://Website Walter Brewer

    Yea but do you realize that adobe can’t even make Flash run right on dual core desktop! I cant even watch a low quality utube video without my cpu peaking like 40%. You also have to remember that Google is a search company that makes it’s money from selling Ads. They can deliver more Ads if everyone is using programs in their cloud. That is why they push “web apps” and “cloud computing” on everyone. Like others have said, Web apps are good for some things but not for everything. There will always be a reason to have both EVEN if web apps can match native app performance.

  • Pingback: Google Maps Navigation goes live in 11 more coutries | News URL()

  • Pingback: A Boy And His Android » Blog Archive » Google Maps Navigation goes live in 11 more countries()

  • Pingback: Game scope()