The mobile web app vs native apps debate is an interesting one. One of the speakers at yesterday’s iOSDevCamp event delivered a good talk on mobile web apps. Here are some of the topics we discussed and debated.
Number of touch-optimized mobile websites: The speaker provided data to show that the number of touch-optimized mobile websites was a multiple of the total number of iPhone and Android apps. He also made the point that mobile websites had more users than native mobile apps.
The data on web-app numbers and user-numbers are interesting, but I think that for many businesses, they are far less important than other factors such as Revenue potential, Discoverability/Distribution and User experience. On these counts, native apps and the app store generally do much better than mobile-web apps.
User Experience: In general, as a practical matter, on the iPhone, native apps offer a far superior user experience to mobile web apps. In almost every instance where the iPhone user has a native app option, they are more likely to use the native app instead of an equivalent web site.
At the talk, the speaker promoted mobile web apps and mentioned that the New York Times app on the iPhone was just a “web application” with a “native shell” and hosted Safari web controls and “web content”, but I pointed out that this assertion was incorrect. The iPhone NYT app is a native iPhone app and (from its V1 days) it seems to have been using standard Cocoa Touch classes such as UITabBarController, UITableViewController etc. Almost every iPhone NYT reader uses the NYT app instead of the NYT website.
Revenue potential: In my opinion, native apps sold through the app store offer a much superior revenue model for mobile apps. Web apps generally find it a lot more difficult to charge users for products/services and ad-funded models for web apps often don’t yield much revenue. As an example, the developer of the iPad Elements app said that their iPad app earned more revenue in one day (through the app store) than what their equivalent web app (PeriodicTable.com) had earned with ads in four years.
Reach: A single mobile web app can generally run on several different types of mobile devices and therefore it will offer a wider reach than any one single native app. So if being accessible from a wide range of devices is very important (or if the appearance of providing equal service to all phones is important), the mobile web may be a better option than building multiple native apps. However, there are caveats with the assumptions of wider reach.
A decade ago, at Microsoft, one of the teams I led worked on the development of mobile apps. Based on that experience, I’d caution against assumptions that a mobile web app tested on one device will run well on all mobile devices. For that matter, cross-browser issues have been a problem on desktop browsers as well.
Is it possible for the web to crush native apps: Anything is possible, but it is unlikely that HTML5 or other mobile web apps will win over native apps in the near-term.
A resource-rich organization can certainly invest inmobile web apps *and* native mobile apps now. That would be a good way to hedge their bets. However, many organizations may not have the resources to invest in web-apps and native apps and will need to decide whether they need to invest on sophisticated mobile web-apps or (one or more) native mobile apps.
Cascade Software Corporation 
On the revenue topic, yes the app store is a much clearer way to get users to pay for you app. However, it seems that this is orthogonal to the question of whether your UI is built as fully native or whether it is just HTML/JS hosted in a web control. You can offer a product that was built using either approach via the app store and continue to have the same revenue opportunities.
That said, IMO there is a more important philosophical issue with putting information behind these native containers as opposed to the web. A good friend in the know recently predicted that (due to all these ‘native’ apps), the web is starting to see a decline in user generated content. We talked about whether this was a result of ‘native’ apps or whether it was a result of people using mobile devices more than PCs and a side effect of input capabilities being a lot more inferior on those. IOW, it’s hard to do a lot of written blog posts via a mobile device. Perhaps we are all shifting towards more consumption and less creation. Anyway, I agree with my friend that this is probably a more important thing to track and manage than the differences in revenue opportunities between native vs. web approaches.
On 1) yes, as you suggest, a hybrid app (native shell with web controls and HTML/JS web content) can overcome some of the limitations of web apps.
However, as noted in the post, native apps are generally going to provide a better user experience. In spite of advances in HTML5, HTML/JS (on any mobile platform) isn’t as rich/powerful/fast as CocoaTouch/Objective-C or Android/Java. For that matter, keep in mind that – even on the desktop – Microsoft chose to implement a Silverlight runtime rather than rely on HTML/JS.
Incidentally, for some UI display in our iPhone apps, we have used UIWebViews to display HTML content. However, the apps are all implemented with Objective-C and I see the app’s HTML content just like I see the app’s string resources.
Most mobile web proponents say that HTML/JS will become much more powerful in the future and that it won’t be plagued by browser incompatibilities similar to what we say with WAP browsers over the past decade (and with all browsers in general). However, for the most part, it seems more sensible to make today’s product decisions based on current tech realities instead of the future “hopes” of web ideologues 🙂
On 2, the philosophical issue you’ve described is interesting and important.
On one hand, the web today suffers from too much junk and so I wouldn’t mourn too much if there is a decline in user-generated content 🙂
On the other hand, and more seriously, I think your friend brings up a good topic. I think that the decline may be related to multiple factors.
In addition to what you’ve described, I’d say that as technology and human behavior keep evolving (for good and bad), people are going to generate and consume content differently. For instance, the popularity of micro-blogging (with Twitter) means that people are going to rely lesser on blogs and RSS feeds. Twitter, Facebook Places (introduced last week), Foursquare etc. result in a lot of user-generated content that wasn’t being generated a few years ago.
Interesting points. I’m intrigued by the native vs. web app debate. Regarding the NYTimes – their native app is well done and definitely streamlines their content for the iPhone. But I think it’s unfair to compare it to viewing NYTimes.com on mobile safari because the NY Times does not serve up mobile-optimized content to mobile safari – they simply serve up their entire website (which is more suitably viewed on a desktop browser). Why did the NYTimes make that choice? Who knows. They don’t sell the iPhone app, so it’s not for revenue reasons. Anyone with insights on that?
Interesting topic, Ram.
Though my sensibilities lie with web apps, I can see that native apps can be are are now, useful, obviously; and HTML5/CSS3/JS/SVG are still evolving from a UI standpoint.
(Tangentially, a BIG problem is developers are still learning about why browser-sniffing is ugly & lazy…while “compatibility-detection” is better to help with browser/device diversity. And Google seems to be browser-sniffing for Opera to block Google Instant functionality, which is even uglier.)
I think as other browser firms try to simplify usage, it’s going to take HUGE leaps of usability to catch natives apps; i.e., Opera adds a “www” in the address bar in Opera Mini/Mobile, but millions of non-tech users still search URLs in search engines, still type “.com” or “www” in address bars, still are being confused, etc… When I’m helping friends/family, I have to hide my cringing…we do a TERRIBLE job of helping users not feel stupid using their browsers.
That’s my guess why native apps are skyrocketing, plus the cool touch capabilities, one-tap install and easier discoverability in the app stores (especially with iPad design possibilities).
This will be THE central debate in the next few years as more varied devices proliferate, I think.
@dadshouse: I’ve actually emailed NYT about their mobile web site, after trying it…it’s quite awesome, but as you said, they don’t redirect mSafari or Opera Mini/Opera Mobile or others. Not sure why, and they didn’t respond to my suggestions.
@Ram: Congrats on your app design, they’re incredible.