To compete favorably with OTTs and device vendors on applications, and be appealing to developers, Telcos firstly need to get a momentous position on application markets. They could appropriately target the “very large and mostly underserved market at the high end of the feature phone category”, where they would benefit from a structural and distinctive advantage. By definition, OTTs don’t own telecom networks, which enable to serve feature phones on a browsing mode. Most of recent handsets run Java, hence have an embedded browser.

Besides, Telcos could deliver services to Smartphones too for non-recurring usages. The point is that downloading client SWs for occasionally used services would be rapidly cumbersome. Therefore Telcos could opportunely expand their footprint on application markets from features phones towards Smartphones, since they would be legitimate in the business as on the other hand they already control the channel to market.

Beyond that, once they will be firmly positioned on the application business, Telcos could leverage other differentiating assets.

Although a future where “the web is the platform” (as Google likes to say) is very likely, it is still in the distant future for two main reasons:

1) Browser replacement cycles are long.

For HTML5 Rich Internet Apps to replace native apps we need to get to a point in time where HTML5 browsers are widely deployed. I have not seen a forecast but I think it is safe to say this won't happen for a few more years because browser replacement cycles are long. As an example on the PC Internet Explorer 6 still enjoys a 20% market share almost 5 years after IE7 was launched. Although mobile replacement cycles are shorter, it's still a slow process that will take some time.

2) Handset manufacturers such as Apple and RIM will continue to push for native applications that take advantage of the unique capabilities of their devices.

In order to maintain differentiation device makers want apps that run better on their devices. Because of this they will continue to expose APIs that are unique to their platforms and will encourage developers to write apps that take advantage of them. You can easily imagine that in an HTML5 world where "the web is the platform,” applications are more likely to offer the same experience across all devices. This tends to commoditize the device makers as it curtails their differentiation.

Case in point: I just attended RIM's developer conference at Mobile World Congress in Barcelona and the number 1 message RIM communicated to developers was: develop your applications to integrate deeply with the RIM platform. RIM has exposed a number of proprietary APIs that enable third party applications to integrate with RIM's own applications and is pushing them hard. RIM is even branding apps that take advantage of such proprietary APIs as "Super Apps." As stated on RIM's own blog: Super Apps are "apps that become a seamless part of the core user experience, an extension of the inbox, calendar, address book and other native apps." It's a good pitch. However it's also a way to ensure that these apps are tied to the RIM platform.

The walled gardens of today's App Stores will be challenged by open web standards and web runtimes in the years to come. HTML5, CSS3, enhanced 3D rendering in browsers etc. will slowly start to loosen the binding between content and devices (and device dependent eco systems). The browser environment will prevail over native OS apps, making room for entirely new players to fight for the best retail position for mobile content.

Start tweaking your recommendation engine today!

What about the other topic on the original post: what role will OPERATOR STORES play in the app store landscape? Do you believe operator stores will reach critical mass? How will operator stores differentiate? Will they be successful on smartphones or only on feature phones?

Thanks for your comment. I agree that fear of malware is a good reason for consumers preferring a trusted and branded store. With more open operating systems on devices, the risk of virus, spam and phishing attacks increases.

Thanks for weighing in the discussion. I do think that the department store and shopping mall models could merge, as long as the brand of the specialty stores in the mall doesn’t get in the way the discovery process and the user experience. The No App Store model is a little different because there would be no intermediary between the developer and the consumer. In the PC world, it would be like buying Quicken directly from Intuit and Photoshop from Adobe. This is a possible scenario, but I think it is unlikely as long as the developer community is highly fragmented.

To add to the discussion, Bango just published a new article that supports Matt’s “No App Store” model: “Life beyond the app store, developers look to go direct” at this link: http://news.bango.com/2010/02/03/life-beyond-the-app-store/ I can see many developers selling direct as a way to supplement their app store sales, but not as a replacement of the app stores.

couldn’t agree more here – those customers that by app’s like to know that what they download and install isn’t going to cause issues with their phone. So a trusted source is key here.

Perhaps the three models are indeed different *angles* into the same future of App Stores. Here's how I see these models converging into a "unified theory":

1. The Dime-A-Dozen view talks about diversity and sophisticated segmentation of apps, i.e. where there is an app store for each lifestyle segment or geographical micro-region. In this instance, app stores are the equivalent of App Channels.

2. The Winner Takes All view talks about a single App Store provider per platform. I see this applying to a single *department store* provider per device or platform, but in general 5-10 mega department stores dominating the landscape. Each department store has multiple shops-in-shop i.e. App Channels.

3. The No-App-Store view talks (in part) about app stores disappearing or becoming transparent and the only value being that of discovering apps. I would agree in that the boundary App Store building blocks (i.e. developer market and discovery channels) will be the ones that remain visible and differentiating, whereas the remaining building blocks (billing, distribution, delivery) will become transparent and offered as commodity blocks by the mega department stores.

Andreas

1. App stores will be “a dime a dozen” (Andreas Constantinou’s original post)

2. App stores will consolidate significantly: “winner-take-all” (this post)

3. App stores will disappear altogether: “no app store future” (Matt Millar’s great contribution above)

I’d love to hear from others as well. Which model do you think is most likely?

Thanks for the elaboration. You make really good points. However I agree with Martin. The “no app store” model you propose where content is scattered throughout the internet (rather than centralized in a store) makes more sense for free internet content than for platform dependent, paid applications. I think consumers will prefer a trusted and branded source for the latter. But only time will tell us for sure.

I do think Apple deserves much more credit than you are implying with your comment. I think the success of the App Store IS the result of Apple’s innovation much more so than the result of consumer familiarity with mobile apps, consumer self selection, and Apple advertising. In particular, Apple:

1. CREATED A MARKET with a device that is great for consuming applications
2. LOWERED THE ENTRY BARRIERS FOR DEVELOPERS with a shorter time to market, no certification fees, and reduced device fragmentation; and
3. CREATED THE APP STORE: a distribution channel for developers to directly reach consumers without intermediaries, with unlimited shelf space, developer set pricing, a generous rev share, and a great user experience.

As you pointed out, many attempts at app stores failed during the “Bi” era (Before the iPhone). But they did not fail because stores were not needed. They failed because of poor design. Although they appeared as “competent attempts” at the time, with hindsight, I think Apple has shown us otherwise.