APIs (Application Programming Interfaces) are becoming a common sight as more and more companies launch them. However, they are often perceived as add-ons to existing services rather than of primary importance. However, recent rapid growth in the number of APIs and more financing in the API infrastructure space (for providers Mashery and Apigee) suggests that there may be something deeper worth looking into. In many cases, it is the APIs that are becoming the main event and not the individual apps they drive.

Mobile applications have become the driving force of transformation on the Web, creating not only whole new categories of software, but also creating new ways of consuming content and accessing services. App sales have also created new revenue sources for app creators and platform owners alike, with $5.5 billion already paid out to App developers by June this year, and Android revenues, while still behind, are also picking up.

While these numbers are impressive, they only tell part of the story. Looking under the covers, the nature of the mobile apps now selling has changed radically since the debut of the app store in 2007. While individual stand-alone apps still do become highly popular, increasingly the real power of apps is in what they connect to from the device.

Today, 77 percent of the current top-50 free and top-50 paid apps connect to backend services of some kind (including game social networks), and only 23 percent were completely stand alone (never calling home to some kind of backend).

Looking deeper, many of the most long-term successful apps are powerful, mobile clients that are part of a multi-channel distribution strategy in which apps on mobile platforms, Web access, and distribution via partners all play a role. The top apps across Free and Paid (based on Apple’s current iTunes App store rankings) currently include apps such as: Pintrest, Spotify, Pandora, Facebook, Netflix, Instagram, WhatsApp Messenger, and Skype.

In each case, the applications are free and provide value by giving users new ways to access backend services. Media apps such as Flipboard similarly draw content from Flipboard’s own servers and from many third party partners.

The same dynamics apply in categories like productivity. While there are heavyweight Apps such as Apple’s iWork suite, the leading apps also include Evernote, Instapaper, Box, and Dropbox, all of which tie into multi-platform strategies. In gaming, there is increasing value in gaming backend services for saves or community interactions to keep people engaged. Companies such as Funsockets are working on specialized sharable gaming backend services.

The number of APIs themselves is also growing rapidly with over 1,000 new APIs being added to the Programmable Web public directory in the last three months. By most estimates, the Programmable Web catalogue represents around 20-30 percent of the total APIs in the world, since many APIs remain single application only, in-house, or for closed sets of partners.

Drilling down, it is very often the case that the apps that sustain long term popularity are part of a wider platform strategy, connecting to backend APIs either provided by the same company as the App or to third parties on the basis of licensing agreements. This enables companies to provide a broader, more consolidated offering, reach across multiple platforms and (for productivity tools in particular) create revenue streams based on recurring subscriptions rather than one time app sales. An obvious additional benefit is that platform subscription revenues are not subject to Apple or Android’s revenue cut for App Store sales (although this is by no means guaranteed long term).

In the recent heated debate about the purchase and shuttering of Sparrow by Google, some of the most interesting discussions centered on the challenges Apps like Sparrow face with a “pay once, use forever” model in today’s app stores. David Barnard’s excellent post describes the economics of a similar app. In a nutshell, app store norms have tended to foster the expectation of low price points and make it very hard for apps to create sustained revenue stream unless they can consistently generate massive growth in new adopters, month on month.

Ultimately with many apps, it is actually the backend service (in Sparrow’s case, email), which is where the sustained value for the user lies. This puts companies that roll out backend platforms, which can keep users coming back over time and potentially access via many different devices, in a much stronger position to capture value in the long term. As the number of mobile OSes and device types increases, consumers, especially of social apps that rely on network effects, increasingly want delivery formats on all platforms and hence require backend API integration.

The proliferation of mobile backend platforms like Stackmob, Parse, Cloudmine, Kinvey, and Appcelerator’s recent acquisition of Cocoafish also underlines this trend, aiming to make it easier to manage the standard user data often needed by apps.

In a survey of 1,000 of the latest APIs to be listed in Programmable Web’s global directory of APIs, while 38 perent still had exclusively non-mobile use cases, 11 percent of the APIs had “mobile only” targets, but 51 percent clearly had both mobile and non-mobile use-cases in mind.

In almost all cases, APIs are being launched as the core of a delivery strategy for the content or service in question. The multi-channel strategy means companies can:

  • Focus on their core value.
  • Deliver it whichever OS platform their customers require it.
  • Open up part of the ecosystem to enable third parties to build some of the clients they do not have resource for.
  • Make it their backend service, which is sticky over time and monetizable.
  • Potentially avoid some of the high costs associated with actual sale via the mobile App Stores.

It doesn’t have to be complex. While operating an API is certainly different from operating a Web property, in many ways it fosters clarity in target audience and the core value being delivered. There’s also no need to go “fully open” with an API and release it to any developer that wants to use it: Often well-managed in house use is a very valuable start. The precise technical approach is not as important as fostering some “API thinking” and mapping out how core functions of a service could be accessed by in house and third party clients.

On the deeper technical side, there are now great libraries in many Web frameworks for exposing APIs and toolsets are evolving. There are also infrastructure providers such as 3scaleMasheryApigee out there that can provide many of the nuts and bolts needed to get an API live and make it succeed.

[Image courtesy SEO Worldwide]