The Dangers of “Sponsored Data”
AT&T this week announced a new program under which providers of online services or apps can negotiate special deals to “sponsor” data usage by AT&T mobile customers, allowing those customers to use the services or apps without it counting towards their bandwidth caps. The argument for this idea is that consumers get some relief from the feeling that the data meter is always running, and online services get a way to overcome possible consumer reluctance to make full use of data-intensive services. But if AT&T’s model were to catch on widely, it would carry serious risks to the Internet’s open and innovative nature.
To be sure, one can imagine scenarios in which a sponsored data option could actually help spur competition and uptake of mobile online services. If consumers are reluctant to try new apps or services due to fear of hitting their data caps, then sponsoring data could provide a way to overcome that barrier and encourage consumers to experiment. In effect, it would provide a way to give out free samples.
If sponsored data arrangements become a more pervasive part of the online ecosystem, however, the risks to innovation will outweigh any potential benefit. Here’s why.
First and foremost, sponsored data arrangements could empower ISPs to pick winners and losers in the online marketplace. The core principle at stake in the debate over Internet neutrality has always been “innovation without permission” – the idea that Internet innovators can roll out products and services to a worldwide customer base without having to get permission from or otherwise negotiate with the various ISPs serving those customers. Online services succeed or fail on their own merits and based on the reactions of end users, not based on the choices or business preferences of ISPs. Sponsored data subverts this principle by giving ISPs a potentially powerful new tool to favor certain services and steer users toward the ISPs’ preferred partners.
This concern might be alleviated to some extent if an ISP offered data sponsorship on generally available, nondiscriminatory terms, but that doesn’t appear to be what AT&T is doing. According to the Washington Post report, AT&T’s initial sponsorship deals were negotiated separately and charge sponsors different prices. So it appears to be all a matter of individual negotiation. Want to make sure your new app doesn’t start off at an immediate competitive disadvantage to larger, more established competitors? You’d have to forget that old mantra of “innovation without permission” – and start negotiating with AT&T and any other ISPs that follow suit.
Second, even if ISPs don’t play favorites in cutting deals, sponsored data could raise costs for new entrants and innovators. If sponsored data is embraced by established online services, would-be competitors may feel they need to pay the ISPs’ fees in order to be competitive. Indeed, if consumers come to expect data sponsorship, even providers of apps that use little bandwidth (and hence carry very little risk of pushing users over their caps) may come to feel that their traffic needs to be marked as “un-metered” to get a fair shake in the marketplace. In this way, sponsored data could raise the costs facing online innovators.
Third, sponsored data can create perverse incentives for ISPs not to raise their data caps. The less subscribers worry about hitting their data caps, the less interested online services will be in shelling out for cap-free treatment. If sponsored data becomes a major source of ISP revenue, ISPs will have a tangible, bottom-line reason to avoid roomier caps that would undercut the need for sponsored data in the first place.
The second sentence of AT&T’s press release compares its sponsored data program with 800 numbers and retailers offering free shipping. But those are flawed analogies. 800 numbers and pre-paid shipping don’t give the phone or shipping provider any way to pick winners and losers; the phone companies and shippers don’t play favorites, so any business can get its own 800 number or pay for shipping on substantially equal terms. Businesses also have multiple shippers to choose from, whereas an app provider hoping to reach AT&T’s mobile subscribers can pre-pay data charges only through AT&T. Finally, the use of 800 numbers and pre-paid shipping doesn’t create perverse incentives akin to the ISPs’ incentive not to increase data caps. So-called “two-sided markets” have their place, but, as CDT recently noted, they’re an awkward fit for Internet access service.
In the Internet context, the idea of sponsored data isn’t completely new. For example, services like Facebook and Google have subsidized access to limited versions of their offerings, especially in countries where many people can’t afford a full data plan. But sponsoring data to introduce underserved populations to online services is different than making data sponsorship a feature of general Internet service for major carriers in the United States. If negotiating deals with an ISP becomes a necessary step in offering successful online services to that ISP’s customers, the result might be just fine with ISPs – but for online innovation, it would be a serious blow.
Of course, AT&T’s announcement doesn’t guarantee that sponsored data arrangements will catch on widely. AT&T announced only three initial service provider customers, and they aren’t the online services that are household names. Moreover, not all ISPs have data caps, making sponsored data irrelevant for their subscribers. But it’s a dangerous road to start going down.