Michael King, director of enterprise strategy at Appcelerator, has a good vantage point from which to observe the ongoing debate over HTML5. Appcelerator has helped developers create over 50,000 mobile apps, including apps native to iOS and Android, as well as HTML5 mobile Web apps. King’s opinion on HTML5 is simple: While native apps have a clear edge in terms of features and user experience, both HTML5 advocates and critics may be missing the point. HTML5 is not necessarily an all-or-nothing proposition.
(For a good overview of what HTML5 is and what it could mean for mobile, we recommend reading our January HTML5 report first.)
Q: It seems like the pendulum periodically swings towards or away from HTML5. What is the sentiment toward HTML5 right now?
You always see these swings. Right now, app developers are definitely swinging away from the Web and away from HTML5. They’re leaning more toward native apps. And that’s because, quite frankly, the HTML5 apps they have built aren’t performing and the HTML5 apps they’re testing aren’t performing. One problem is that there’s a fair amount of Web browser fragmentation. A fair amount of interpretation of the standard remains to be done, and given its immaturity, HTML5 ends up adapting to the lowest common denominator among browsers. HTML5 is a great story for the guys who want to build the apps. It’s a lousy story for the guys who have to use them. If you believe that apps are all about the user experience, which is kind of where we stand, you can’t get what you want from HTML5 right now.
Q: Detractors and advocates of HTML5 often make it seem as if it’s all or nothing. Is that really the case?
It’s not necessarily an all or nothing type thing. That’s not an entirely accurate characterization. You have enterprises, for example, that are making decisions on a single app. That’s one kind of decision. It’s a different decision if they have to make 19 apps. You look at it as one core tenet of your mobile strategy. How much HTML5 will I use? We have some clients who are building 100 to 200 apps in a year, and some of those apps are going to be native and some of them are going to be HTML5. It depends what the apps are for.
We have a construct which is called slope of interactivity. The higher up the slope you go, the more interactive the app. Your requirements for a native functionality grow as you move farther up the slope. Something like Netflix video consumption isn’t very interactive — apps like that are a great place to use HTML5.
Q: Aren’t there tools out there offering fixes to HTML5 that will help bridge the gap as HTML5 functionalities catch up to those offered by native apps?
If you look at an open source solution like PhoneGap, they say: Develop in HTML5, and we’ll provide these fixes — meanwhile HTML5 will get there. The expectation is that eventually the Web will understand the phone and HTML5 will get there. But there’s a fundamental flaw with this kind of thinking. This is not about mobilizing the Web or putting a desktop experience on mobile. Mobile represents a fundamental shift in how people interact with data and content. You can use stopgaps, or you can realise this is a paradigm shift and put yourself in the era of mobility.
Mobile apps are fundamentally about how you interact with data. Mobile apps are contextual, they’re about delivering the right piece of information in my moment of need. If I’m waiting for a train, I don’t need to look at the train schedule, I need to know when the train is going to get there. If I’m at the airport, push my gate notification. HTML5 is not good at this kind of interaction, it’s better at content consumption, like Netflix’s HTML5 app.
Q: What about hybrid apps? Do they combine the features of HTML5 and native apps?
I see two types of hybrid apps. There are hybrid apps in which you have any number of native app functions that can be called out though a bridge lay and just use a Web view for limited information. And those that have a native wrapper that they put around HTML5, and it’s making callouts from this wrapper.
The latter type has limited functionality, maybe 12 to 25 feature callouts. It’s a one-size-fits-all experience. If mobile applications are about performance and experience, then you won’t be able to capture the users if it looks and feels the same everywhere. If you want a well-performing app you have to have more native code.
Q: What are some HTML5 apps that have done well?
The Netflix app is a wonderful example of pure HTML5. The original LinkedIn was pretty decent and that was pure HTML5. Though now LinkedIn has been moving away from that.
Q: So the bottom line is that developers and publishers need to look at both options and see what approach is best for them?
It’s very much about being able to ask yourself what functions of the app are required to be native and what functions of an app can I render in an HTML5 Web view. Looking at it like that, it does sort of give you an infinite spectrum of possibilities. It’s not about locking into a specific functionality and a specific architecture. You have to look at both choices. Mobile banking, for example. Not every user’s going to want all the features. They might get an email and check their balance via their mobile browser once a month. They’re not going to want to download an app for that. But some users will want a full-featured app. You have to service both of those customers. We see it from the perspective of — you’re going to do both.
Q: What about the future? Will the pendulum swing in a different direction?
For the foreseeable future I don’t see this changing. I don’t see HTML5 all of a sudden racing ahead. And I don’t see the Googles and Apples and Microsofts of the world dumping all their unique features tied to their platforms and going HTML5. These companies do not really want their handsets to become Internet terminals, and run the same apps. They want to keep the differentiation between their platforms. You differentiate based on user experience. That’s how they are going to sell more devices. Finally, monetization is an issue. There’s no centralized marketplace, no Google Play, for HTML5. No one ever said HTML5 was going to help you monetise.
Business Insider Emails & Alerts
Site highlights each day to your inbox.