It has been well on nine months since I last posted, and I am nothing short of disgusted with myself – a guy who promotes blogging regularly, even if it is just a short post. In a classic case of “do as I say, not as I do,” I got so lost in a major project – unexpectedly so, may I add – that time for blogging disappeared faster than newly released iPads do from store shelves.
The project took me back to my online product development roots, and allowed me to build an incredible analytics system for mobile devices. I figured I knew how to build algorithms and real-time search engines, so it might be good for me to know how to build the analytics system that measures the results and feed them back into the behavior of the engine.
I am now officially hooked on mobile as part of the repertoire. Mobile today reminds me of the web in 1996 – 1998 – it is the wild west of technology markets. The rise of smartphones and tablets has changed the entire dynamic of how people interact with information. Everyone is experimenting, and their are thousands of small firms making bets on various approaches to mobile. Evolution is fast and furious, with winners and losers coming and going almost overnight as market changes and technology enhancements literally create disruptive change in the ecosystem.
One ongoing technology argument is HTML5 versus native apps and whether HTML5′s advantages of cross-platform compatibility, channel freedom (not being dependent on the app stores), enhanced discoverability due to better search engines, large base of developers, better analytics tools, and substantially lower cost of development will cause it to overtake native app development as the platform of choice for next-generation mobile apps.
I was at the AppNation conference last week and Trip Hawkins – someone who knows about as much about mobile gaming as anyone – weighed in.
“The browser will beat the app store,” he argued. “It’s more convenient, it’s driven by search and it’s more viral for consumers. It doesn’t matter what device your friend has… All technologies start as silos, but get 100 times bigger when they become inter-operable; think about roads or text messaging.”
I agree with him completely – ubiquity trumps functionality every time, and ultimately the ubiquitous platform surpasses the proprietary platform because more people invest in evolving it to do what they need it to do. Think open source and the “community of the commons” which helped evolve platforms like Java, Linux, Twitter, and Facebook. Or even more fundamentally and applicable to the HTML5 discussion, think the original Apple versus Microsoft Windows. Windows ultimately dominated – and still dominates – Apple as a platform because Microsoft early on allowed its OS to be licensed to any hw manufacturer, extending the franchise and opening the market to an overwhelming level of app development investment that to this day dwarfs what Apple can offer.
This is the economic law that Brian Arthur has described and quantified in his evolving work on Complexity Economics and something we all understand intuitively nowadays. That is, the value of the network (think technology network) increases as it becomes more ubiquitous, thus drawing more investment into it which only further increases its value. It is a virtuous cycle, and one reason becoming the market share leader in a new technology is so critical.
Apple has learned a few things over the years, and with the iPhone and iPad they have quickly moved to gain dominant share and the largest developer community. They are evolving their tools for developers and their OS faster in order to keep their developer community growing and provide more opportunities for developers to make money on the platform. It gives Apple a tremendous lead and advantage at this point against any open platform alternative.
But developing native apps is, relative to HTML5, time consuming and expensive. Unless you are in an industry like gaming where the quality of immersive graphics and performance is critical, HTML5 offers a cost and time to market advantage that provides a lower barrier to entry into the mobile market.
Then there is also an arrogance to Apple that is driving developers to consider ways to bypass the platform. Consider:
- The iTunes store is a bottleneck. It is a single place to sell, and it gives Apple tremendous power over who succeeds and who does not. And Apple has been more than willing to display its dominance to developers and the community at large.
- There is a relatively and unpredictably long approval process to get an app live, and often apps are rejected for issues that were not apparent in advance.
- Apple changes the rules continuously and in many people’s minds, arbitrarily. Take the recent paid download spat with Tapjoy and others. In a single dictum, Apple basically removed a major monetization mechanism for developers.
- You cannot tell when a push notification actually goes out through Apple. So you pay someone like Urban Airship to send the token to Apple, but then about 1/3 of the time the pushes never arrive. Apple is a “black box” on this and won’t tell you why (although they say they are working to fix this). Now if that isn’t arrogance, I don’t know what is.
- Apple keeps track of everything developers’ users do on their platform, but do not make the data easily accessible to allow developers to improve their products and optimize their revenues.
All of this give Google and Windows 7 a potential opening to exploit HTML5′s advantages, and they will. Google is already doing so, although they are to a certain extent hedging their bets right now. And Facebook pushing HTML5 apps gives the platform momentum. Eric Schmidt was at Sun with me when Java came to market, and knows first-hand how to leverage an open platform to disrupt an entrenched proprietary competitor. Voila Android. And very soon, Android devices will have a larger installed base than iOS, if it doesn’t already.
The weakness Android has in the app market is fragmentation. In my limited experience looking at the data, there have got to be at least 30 variations of Android out there. When you look at the ROI for developing on Android, it only pays to develop for one or two versions – and this sentiment has been documented extensively in magazines like Fortune .
But that argues strongly for using HTML5 as a development platform. It allows developers to undercut Apple’s power position in the industry and develop for all Android phones.
HTML5 thus provides the opportunity to develop faster, cheaper, with a wider audience to sell to. And Apple is more than providing the motivation, just like Microsoft did for Java. As a rule, developers are a libertarian bunch who hate being dictated to by anyone, and they are getting increasingly upset with Apple.
The only question is how fast this transition will occur, and that remains to be seen. I think ABI Research is right when they say that native app development will peak in 2013 – that just feels right to me given my experience with numerous technology adoption cycles. But any prediction like that is fraught with peril. On the one hand, some killer mobile app built on HTML5 can come along tomorrow and change user behavior to the point where the browser becomes a natural first point of entry into a mobile device. On the other, Apple could change its modus operandi and salve the wounds of upset developers.
So while it is a matter of time, developers for now will have to make a bet on when HTML5′s time will come.