I’ve been thinking a lot lately about the merits of mobile native development compared to mobile web development. Native mobile development is so complex and fraught with so many pitfalls, and that situation doesn’t look like it’s changing much, despite the advances many handset manufacturers and platform providers trumpet. Myriad technical issues remain, while the difficulty in establishing a business model persists.
Obviously this isn’t a zero-sum game; there are plenty of instances where native apps make a lot more sense than web apps or services (or are the only way to tackle a problem). But are those instances becoming more rare? And will the best mobile devices in the future — in terms of development platforms — just be the ones with the best browser?
Michael Mace beat me to the punch with an excellent post today, asserting that “The business of making native apps for mobile devices is dying, crushed by a fragmented market and restrictive business practices.” The general gist is that native development is a real pain in the ass — and the business model is so broken that the rewards for those who undertake the task aren’t that great anyway.
Mike Rowehl and Dean Bubley both have good responses on their own blogs, and both echo the horses for courses point, while illustrating some of the problems with the native app business environment — in particular, Dean highlights the difficulties of distribution, given the lack of a viral mechanism to spread apps as well as the resistance of most people to install applications.
What got me thinking about all of this was the Nokia N82 I picked up recently. I’ve been pretty happy with it, as the hardware features are great, and the S60 software platform is improving (for one thing, it’s gotten a much-needed speed boost). But it still has a lot of faults from a usability angle: it’s just way too damn complex. Getting all of these great features like Wi-Fi set up are a challenge, even for a seasoned mobile tinkerer like yours truly. I can’t imagine handing one of these to a normob and asking them to get the Wi-Fi working. Furthermore, I spent a good deal of time installing all my usual apps when I got the phone. It’s hard to see an average user spending so much time getting things all set up. In comparison, opening up a web browser and navigating to a page is much simpler (though it remains far too difficult on many browsers).
Mobile web development isn’t without technical issues of its own, fragmentation among all the different browsers paramount among them. And if you’re looking to reach hardware or other functionality of a device, web apps don’t always offer much help (though this should be changing). There are plenty of cases in which native apps still make sense and can thrive (Dean runs through a bunch of them in his post), but I think the confluence of the business and technical environments (and I guess you could say the telecom environment, too), will push many developers towards the mobile web instead of native apps.
Strange as WML / WMLScript technologies seemed at the time to all the HTML writers out there, the point of that exercise was to do what current mobile webpages cannot: bring native interactivity to the phone. Early WAP technologies were the bastard child of Apple HyperCard and HTML, and useful applets had indeed been made for years using HyperCard. WMLScript was made so that actions on a WML Deck could be embedded into the soft-key menus like first class native actions.
It was a terrible fit to the developer market when marketed as Mobile Web when every Web designer expected HTML-lite, and the gutlessness of OMA’s decision to not specify hard form factors in their profiles then killed it because nobody could really design a WML Deck application if they didn’t know how many soft keys they would actually get to work with on what kind of screen. But it was a real attempt to create ‘native like’ applications to be downloaded and browsed to over Port 80. It’s just that nobody really ‘got’ that that was the model, including, after the first attempt, OMA itself.
WiFi is a bitch to set up, therefore the web is in and mobile apps are dead. Irrefutable
Though with all the caveats of the last paragraph, the central argument isn’t without merit.
I think perhaps a better way to present it would be without the “death knell” attention grabbing headlines though: what is being said by a number of commentators is that installed apps are too awkward to become general purpose and most people didn’t want them anyway (so rather then being dead, they were always just niche), but it’s possible that more people might use mobile web-based apps instead, probably in a few years once operators become cheap dumb pipes and various other barriers are overcome.
Basically, an exact mirror of what is happening on the desktop but on smaller more awkward screens; the only sense in which mobile apps are dead is that people expected them to take off like the early PCs and then die away eventually outside of their niches, whereas actually they never really took off at all in a mainstream sense because of collective idiocy among the operators and manufacturers.
Native Apps might be great for the techie or experienced mobile user. But for normobs, it just isn’t a viable consumer solution. Normobs will not, for the most part, download anything onto their phones. They can’t be bothered by it; they will take what the phone has pre-loaded on it; they have heard the horror stories of downloading viruses; they don’t understand how to do it; and once again, they simply don’t want to be bothered by it.
If it solves an incredible problem, then possibly they will get help in setting one up. But that is the minority. Most native apps are “nice-to-haves” and not “must-haves”. If it isn’t a “must-have” it won’t make it to the masses. Remember, consumers simply want solutions to their problems, not complex technology that may help out here or there.
However, often there is a fantastic market for extremely niche apps that have found a way to monetize by only being used by a select group. Therefore, the mobile browser may win the war but there will be many battles continually won by native apps…
[…] The Telco2 blog is noting similar trends, pointing to a Nokia-targeted developer that shortened its app developer cycle to three weeks by using Web-based development styles. In fact, the whole idea of Web versus native mobile apps seems to be making the rounds in the blogosphere (see: Mobile Applications, RIP and Sounding the Death Knell for Native Mobile Apps). […]
[…] the death, to be replaced by a Web-based future? Former Palm and Apple exec Michael Mace thinks so; Carlo Longino agrees. The argument is that the diversity of possible platforms, the difficulty telcos […]