The Antitrust of Internet Platforms and Open-Source Innovation: Google, Twitter, Facebook, Apple (Part I)
As I noted in my prior post, the dominant antitrust issue of the times may be the limits on companies — who create value by developing platforms that are open to others — to rein in or step back from that openness as they flush out their business models. This very openness is at the heart of much internet innovation, but it also leads to fact patterns that superficially fit templates for antitrust claims. This post will explore the characteristics of today’s internet business models that make them inherently susceptible to antitrust. Forthcoming posts will explore the consequences of scrutinizing internet platforms in this way.
Innovation in platform creation is open-source, happening through collaboration of the platform and partners, rather than closed, happening under a single roof by the platform itself. The openness sets up a common story arc that includes some or all of the following elements:
- from the very early stages, platforms are open, inviting and relying on heavy interconnection, and this openness and interconnection is central to the platform innovation itself;
- independent partners, utilizing the open platform, are primary drivers of platform innovation along with the platform creator itself;
- as platform owners seek out a sustainable business model, platform and partners potentially butt heads as they grow to compete in certain areas;
- disputes between platforms and partners arise when the platforms adjust their policies, pulling back somewhat from openness and interconnection or expanding in new directions, or when others demand access or access on specific terms.
The open source model of innovation is inherently more susceptible to antitrust, particularly in the early stages of the platform, than the closed source model, because open source means new economic markets have been created; independent companies, who can also be competitors to the platform, have developed; and acts of the platform can disadvantage these partner-competitors.
New Business Models and Platform Innovation
The first characteristic of new business models is their inherent openness. This is marked by companies opening their ecosystems before they are even fully envisioned or developed, utilizing partners to help build and shape the platform through mechanisms like open APIs providing additional features, services, etc. built on top of or as extensions of the platform. The internet’s power – including that of many of its iconic success stories– is built on this openness.
For example, Facebook and Twitter expose APIs to allow developers to build applications and businesses on top of their platforms to increase engagement with their own platforms. For Facebook, Zynga’s games have been a big driver of usage. For Twitter, developers have built interfaces such as client software or url shortening services to give users desirable ways to post on or read tweets off the service. Apple’s AppStore harnesses the power of developers to build applications that drive sales of iPhone and iPads.
The second characteristic, also inherent in new economy markets, is business models that take shape over time. Subsequent discovery and refinement of a business model often requires that the region occupied by the core ecosystem broadens or narrows as the business model finds its footing. This can mean that the platform finds itself in competition with its partners eventually, as it finds a revenue opportunity or expands to take over a core function of the platform.
There is then a recurrent pattern — initial openness followed by adjustments or “stepping back” as business models evolve. New economy companies have numerous reasons why they must then “manage” the openness. For example, Facebook or Twitter or Apple have an interest in preventing the tarnishing of their platform by malicious apps or in controlling the look and feel by controlling the client software.
Even more importantly, the need to manage openness comes up in the all-important search for a sustainable revenue model. The companies must eventually monetize their platforms, gaining a return for the investment in creating a platform that brings an audience or customer base to the partners. As an example, Facebook takes a revenue share from its apps in the form of Credits and Apple takes a share of AppStore purchases. Or eventually, the platform may discover that it makes sense to monetize the business directly, so for example, Google may want to be more of a travel player as it realizes that its search engine is being used for travel.
Twitter, in particular, recently has had to face these issues, because it was so open to its developers early on, having to prioritize and cede some core functionality to independent developers, later realizing that it must it control more of its own ecosystem to build the service it wants. As one of its employees noted:
“In the early days, all the clients except Twitter.com were built out by ecosystem companies, mainly because Twitter was so focused on keeping the lights on, but we learned that in order for us to really grow, we had to start taking over that core experience.”
Antitrust and Platform Innovation
These recurring aspects of internet business models – openness to partners, evolving business models that find platforms in competition with their partners, and stepping back from the openness to implement those business models– make internet platforms vulnerable to antitrust. When there is the inevitable dispute between a platform and a partner (whether someone who has been granted access or wants access), this recurring fact pattern means disputes where expectations of partners are unsettled can be framed as antitrust matters. More or less, all the recent new economy antitrust matters whether involving Twitter (and issues with the developer community), Google (and issues with those challenging its rankings), Apple (issues around the AppStore), etc., can be seen through this lens.
To understand why, let’s step back for basic antitrust doctrine and how it corresponds to these recurring aspects of internet business models.
For an antitrust claim, one needs an impact on competition on a market. Having given access to others to build services, the platforms have also created new and separate economic markets. Ironically enough, except for the prior openness, games on the Facebook platform or Apps for iPhones would not even be an independent economic market, but perhaps would have been characterized as a feature of the platform that the platform itself provided.
In addition to separate economic markets, the platform and partner are potentially competitors to provide a specific service on a platform. Where business models are so fluid, and there is trial and error and discovery as a platform figures out how it can make money or provide the best user experience, there is the potential for a situation where the platform merges into the same competitive lanes as someone who has built or promoted a competitive business using the platform. So, in the starkest case, a platform may realize that it should be doing something directly what a partner is doing, e.g., it should sell travel and not just direct its users to other sites, it should market its own games instead of relying on others to do so, or it should create its own interfaces. So, the platform becomes a competitor to certain partners.
The notion that (1) there are new and separate economic markets in which the (2) platform and partner are competitors has important antitrust consequences. Typically, if one invests in and develops some asset, it is under no obligation to allow others to access it to build their own businesses. While the partners couldn’t have forced the platform to open up these markets, in the same way that I cannot force Netflix to carry my home videos, it is harder to put the genie back in the bottle once it has been opened. Now actions in which the platforms manage their openness that disadvantage these independent competitors where there is a competitive business owned by the platform have the possibility of antitrust implication, as disgruntled competitors try to frame the issue of a “monopolist” of the platform extending its market power to the related market (that only exists because of the openness) in which the disgruntled competitor operated. Whether those claims will hold up at the end of the day is not a clear-cut, black and white matter, but the important point is that companies are under at least some threat that they must maintain openness they have granted or alternatively have openness forced upon them.
The next post will explore the important question: whether there are legitimate antitrust worries or this is an unfortunate situation in which antitrust templates happen to superficially fit onto features of internet business models.