Apps Are Not Killing the Web
Forrester Research is an oft-cited source by businesses when making decisions or declarations about trends and technologies. In many circles Forrester is something of a de facto standard for analysis. As such I fully expect to start dealing with a recent statement from its CEO claiming that the web is dead when I sit down to talk with clients.
On Thursday morning at the DeSilva + Phillips Media Dealmakers Summit, George F. Colony, CEO of Forrester Research was on a panel discussing the future of media in light of tablets and e-readers. Expanding on an answer to a question he fielded, Colony said,
We think the Web is dead.
When he says we, he means the folks at Forrester. Back in October another Forrester staffer was quoted as writing that
the golden age of the Web is coming to an end. This was in an article by The New York Times covering the hub-bub about Wired Magazine’s over-hyped death of the web article. Wired’s article relied heavily, but not totally, on a graph showing the decline of the web as we know it, but Boing Boing quickly refactored that graph with a more accurate visualization using the same numbers. You might recall that I wrote up my own response to Wired’s argument in my post Enough about the Death of the Web
In May of 2009 Forrester also claimed that the smartphone was dead. To be fair, the full title of the study was The Smartphone Is Dead: Long Live Smart Phones and Smart Gadgets, showing that Forrester didn’t really think the concept of the smart phone was dead, but that it was no longer worth breaking into its own category given the ubiquity of capable phones and devices. That may very well be the logic the Forrester CEO was using in his comments, although I don’t think so.
The web is not dying (again). If anything, the advent of tablet and tablet-like devices coupled with support for HTML5 and CSS3 (really just CSS3) in the browsers that are coming on those devices (Webkit-powered Safari and Chrome on iPads, iPhones, and Android devices) is going to ensure that the web will be around for quite a while longer. The Webkit engine (along with the mobile version of Opera that many are grabbing) does a good job of supporting the newest still-in-development standards, creating opportunities for far more interactivity and style than could be achieved in browsers targeting just CSS2.
Angry Birds is a great example of an app that you cannot replicate as easily in HTML/CSS, if at all. But so many other apps are geared toward media, such as eReaders and photo sharing utilities, that delivering much of that content through a browser is a more cost-effective approach. For example, an app like Picplz has to be built separately for both iPhone and Android devices in order to use the cameras built into each. The method you use to browse your photo, profile, and the photos of others, however, is delivered through an embedded browser. This allows the app developers to focus on the features unique to each device while the universal elements are maintained back on their web server, removing the need to push an updated app to users for each minor tweak.
Soon we can expect that an app delivering content of any sort will really be a wrapper for a web browser, handling just the interaction with the hardware and operating system that is necessary for things like user validation, preferences, and so on. This will reduce the cost of app development as the heavy lifting is done via the web server and CSS3 (HTML5 if you’ve bought into the hype). You can expect to see RIM and Microsoft move to catch up by deploying more capable browsers to lower the bar for app developers to deploy to Blackberry and Windows Mobile devices. This closes the gap in available apps for each device, making them more appealing to more users.
This approach also supports users who don’t have these new devices, allowing your typical desktop user with a capable browser to access the same content, even if he/she uses a mouse instead of a finger. A good example of this is the Marvel Comics comic book reader for Chrome. Originally built for the iPad, and supporting a swipe interface, viewing it in Chrome gives you a similar experience but with clicks instead of a more tactile experience.
The new model is build it once, deploy it across the web and your apps. Except it’s not a new model. It just makes sense.
As I was wrapping up this post I stumbled across this article from The Nieman Journalism Lab, The Newsonomics of apps and HTML5. I think that, even though it takes a different path, it comes to similar conclusions as I do.
Update: February 10, 2011
Go read Robert Scoble’s take on the new HP TouchPad. If he’s right and it takes off, building web-based apps seems that much more like a good idea now, doesn’t it? Supporting compiled apps for each of iOS, Android, WinMo, RIM and WebOS seems far less compelling to me.
Update: February 11, 2011
John Dowdell explains the concepts a bit more succinctly over at his Adobe blog: Blends of native and global. It’s pretty much the same argument I make — some bits will have to be developed to run on the device, some bits can be best delivered via the web. Automatically excluding either of those options isn’t a sound business or technical decision.