As noted in my last post, there is an interesting battle for developer and user mind share looming between mobile Apps and mobile Sites (mobile optimized web services that are accessed via a browser). There may be something to learn from the recent comparable historical analog we have seen unfold in the wired PC Internet.
Remember that for the first 20-ish years of the PC industry, there were only applications. Recall that Microsoft held a choke-point in the PC kingdom, and made a huge fraction of the profits in the PC application category. Remember when no one would start, or fund, a PC application business because of Microsoft’s demonstrated ability to subsume that company’s application within some existing Microsoft application? Recall how Microsoft, in their rush to catch and crush Netscape, installed the ubiquitous and essentially choke-point-free browser platform. (In an amusing bit of historical irony, this is strikingly similar to the rushed PC effort at IBM in the early 1980s, which installed Microsoft and Intel into controlling positions in the PC industry.) That new, ubiquitous, and choke-point-free browser platform is what allowed the Internet to flourish. Look at the companies/services that have arisen to date in the wired Internet, along a couple of key attributes.
One key take-away from the table above: while both Apps and Sites existed in the wired Internet, the overwhelming majority of value created in the wired Internet came from the Sites end of the Apps-to-Sites continuum (there is no really clean line between the two), and only a tiny fraction came from Internet connected Apps. Where on the Apps-to-Sites continuum might we expect the bulk of the value creation to happen in the mobile Internet? There seem to be four important differences between the wired and mobile Internet when addressing this question.
(1) OS Diversity. In the wired Internet there were relatively few underlying operating systems that an application developer had to build to if they wanted 80% of all potential users to be able to use their application. Even still, exploiting the installed base of browsers made for much more value creation for Sites over Apps in the wired Internet. The OS diversity problem is at least an order of magnitude more difficult in the mobile world, and while there is likely to be a reduction in the number of mobile OS’s out there, each new release is another version to port/test for an App developer. This would tip the scales even closer to the Sites end of the Apps-to-Sites value creation continuum.
(2) Hardware capability diversity. In the wired Internet, there were only modest differences in the functionality differences between PCs and only a limited amount of non-PC functionality got subsumed into the PC. This led to a fairly uniform set of capabilities for the browser to support. The mobile phone has been subsuming a lot of other functional bits for a long time now (see the figure below). Sure, HTML5, plus something like PhoneGap, allows the browser to be a fairly ubiquitous pre-installed platform, but if one wants to take advantage of the cutting edge hardware capabilities, one almost always has to use native applications to do so. This would tip the scales toward the Apps end of the Apps-to-Sties value creation continuum, especially for point offerings that took advantage of a specific hardware functionality; it would only be a small move along the continuum toward Apps for more offerings that weren’t tied to a specific hardware capability.
Where one needs access to specific hardware capabilities is one of the few situations where a mobile App-based approach should beat out a mobile Site-based approach.
(3) Browser Diversity. While HTLM5, plus something like PhoneGap, allows the browser to be a powerful pre-installed platform, it hasn’t happened yet, and the diversity of interests among the various handset vendors and mobile operators may prevent it from happening as quickly or as ubiquitously as we hope it will (especially without some impending disaster to force it to happen). Recall that the early days of the wired Internet explosion had only a few versions of Netscape and Internet Explorer with the overwhelming majority of the market share. You would have to make a long list of mobile browser releases to get to an equivalent market share in the mobile Internet. And, the early wired Internet browsers were much simpler, which allowed for easy and fast development. This would tip the scales noticeably to the Apps end of the Apps-to-Sites end of value creation continuum.
(4) Head start. Since the iPhone App Store launched in July of 2008, and Android and BlackBerry followed suit within a year, mobile Apps have become the key way that people think about the mobile Internet. The existing App-oriented momentum and collective mindset can be a powerful thing–even though many of the “mobile Apps” out there are really just glorified bookmarks for mobile Sites. The real question is whether or not the currently shipping smartphone browsers are “good enough” to meet the needs of the most popular offerings, and this is likely to vary by category of offering. For example, for real time games, the need for taking advantage of the specific capabilities of each hardware platform would seem to indicate that the balance will be heavily tilted toward Apps. But, for turn based games, which are much less likely to need to take advantage of hardware specific capabilities, Sites would seem to hold the advantage. Generally, this would tip the scales noticeably to the Apps end of the Apps-to-Sites end of the value creation continuum.
Bottom Line. From a value creation perspective, the wired Internet precedent is tipped massively toward the Sites end of the Apps-to-Sites continuum. But, in the mobile Internet, there is tremendous momentum behind mobile Apps; for many categories of offerings, it will be necessary to use Apps to take full advantage of the hardware capabilities; and, the mobile Internet lacks the level of uniformity in browsers that the wired Internet enjoyed in its early days. These lead one to believe that for certain categories of offerings (such as real time games, things needing to access/control the underlying hardware, etc.), the value creation will be moved very much toward the App end of the Apps-to-Sites continuum. But, for all the other offerings, most of the value creation should still be on the Sites end of the Apps-to-Sites continuum. And, since most companies are started by us Engineer-types, too many will believe that they really need to take advantage of the underlying hardware capabilities, and chose an App-based approach, when they would have been better served with a Site-based approach.