The Unwelcome Return of Platform Dependencies

Editor’s Note: The following guest post is written by a Silicon Valley CEO. Frank Dupree is a pen name

In the late 1990s, the rise of the browser was supposed to usher in an era of unprecedented opportunity for startups. A great part of that increased opportunity came as a result of the significant reduction in platform dependencies. No longer did the users’ operating system dictate their access to services or information. Even a behemoth like Microsoft was fighting hand-to-hand combat with small startups for the first time in decades. Fast forward ten years, and it’s 1985 all over again.

But even as the risks of dependencies become better understood by startups and investors, the ascent of Facebook and Twitter seem to point to an ever increasing number of startups with significant business dependencies. Recent changes to both Facebook and Twitter show that neither startups nor their investors can assume much when it comes to support for a given API in the future. Today, even the OS seems subject to dramatic shifts in record time. One only need look to Apple’s iOS to see how dramatic and unpredictable developments can change the landscape for startups, customers and investors.

Today, most startups build with significant external platform dependencies, whether it is Facebook, iOS, Google Apps or Twitter. There are a few types of dependencies. A simple distinction might be to call a dependency on a platform symbiotic or parasitic. Symbiotic dependencies are those for which both sides agree to terms of the dependency and for which both sides seem to derive a benefit. Developers on Facebook’s platform, for example. The most successful here being Zynga, which grew completely and—probably for the team and its investors—nervously within the Facebook ecosystem.

But some dependencies are parasitic. As such, the problem is considerably more sticky when a startup forms an unwelcome dependency. Consider Meebo’s initial products which violated IM network terms of service by “hacking” into the major IM networks. Meebo has had to re-invent itself as a website check-in and sharing platform and at considerable cost (the startup has just announced its most recent round of financing taking it to over $70M in total funding).

In fact, a great portion of the “aggregation” genre of startups have troubling dependencies on larger players who often consider their products either a violation of service terms or simply the next feature on the product roadmap. Take the social media aggregation tools like TweetDeck and Seesmic. It has become clear that even players like Twitter or Facebook who offer generous API access can inadvertently step on the toes of startups in its ecosystem with a new product release. For Seesmic, this appears to mean a third pivot for the startup. The company started as a video commenting platform for blogs, moved into the desktop client social aggregation business and with Twitter’s recent redesign jumped into the plugin (aka, longtail) and enterprise user space (see recent partnership with and plugin announcements). So for early platforms like Twitter, players like Seesmic and TweetDeck exist in a grey space between symbiotic and parasitic for the platform.

Then there are the 100% parasitic players, including companies like Rapleaf. Rapleaf has a dubious reputation because it built its business on the aggressive and surreptitious collection of social network data and matched it to email addresses. Recently both Facebook and Linkedin threatened legal recourse unless the company ceased crawling and scraping their user data and expunged all existing data collected from their networks maintained on over 650 million email addresses. As a result, two weeks ago Rapleaf agreed to stop providing its customers with information from both networks. That can’t be good for business.

But it is likely even worse for the startups who had based much if not all of their products on data they were buying from Rapleaf, sort of a double dependency. Take Gist, Etacts and Rapportive. All three of these startups were dependent on Rapleaf to present social network information (photos, titles, updates, etc.) alongside emails, mostly Gmail. When their Rapleaf data stopped providing Facebook and Linkedin data, the startups were left only with what appeared to be that which they had cached prior to the cut-off this month. And so it may be no surprise that within just a couple of weeks of the Rapleaf changes Gist is reportedly in talks to sell itself to RIM and Etacts has announced its sale to (rumored for $6 million). One wonders if the acquiring companies fully understood the dependencies these startups had to Rapleaf. (Prediction: Rapportive also will be forced to sell very soon).

There is an alternative to parasitic access to a platform, but it can be expensive. You can build your own data from scratch or get access to it through painstaking partnerships if it is critical to your business. You might not grow as fast as you otherwise would, but at least you can control your own destiny. Even for startups that are successful in building their business within a platform dependency, such as Zynga, one of the most important things they must do as they grow is to mitigate their dependency on the platform.

For startups and investors, however, the lure and benefits of developing on other’s platforms with or without permission may be too great to avoid. When it comes to acquiring those critical first users, it takes an Odysseian-level of cunning (or foolishness) to avoid the siren’s song of platforms like Facebook. Building on Facebook reduces both the cost of development and provides direct access to a massive user base. And what are the other options? Even if you put aside the costs of going it alone, a significant portion of startups will be faced with taking on significant dependencies for their success. How many mobile OSes can there be? Forget startups, after iOS and Android, even the future of big players like Microsoft, Nokia and Palm seems uncertain in the mobile OS space.

So in the future it will likely be that the most successful startups will be those which are best able to navigate the minefield of platform dependencies. And while the benefits of platforms like iOS, Facebook and Twitter are significant to reduce development costs and increase access to customers, one can’t help but lament the squashing of the promise we got with our first browsers: a world where startups and the behemoths of industry alike fought on more or less equal terms. But increasingly it is once again becoming a platform world. And the companies who control the platforms, control the profits.

Photo credit: Flickr/uub