The problems with app-based billing for data
AT&T stole the spotlight at Mobile World Congress in Barcelona this week when executive John Donovan told the Wall Street Journal the carrier is considering charging app developers for the data that subscribers use in their apps. While Donovan didn’t disclose many details, it appears developers could voluntarily pay the cost of transmitting content through their apps and that data wouldn’t count against the subscriber’s monthly allotment. The idea, Donovan said, is to find ways to create new revenue streams and have content providers share the cost of disseminating their stuff.
It is an interesting, innovative idea. But it is already being derided as a boondoggle and a mad, terrible plan that is bad for developers. More importantly, though, I don’t see how it is workable.
A solution accompanied by more problems
The proposed plan might seem like a natural fit for siloed apps that don’t integrate any other app or service: Watching a 30-minute video on Netflix, for instance, should be easy enough to monitor to determine how much to charge the developer. But many of the most popular apps are built specifically to work with other apps. Whom should AT&T charge, for instance, when a user views an Instagram photo through Twitter? Or watches a YouTube video posted on a Facebook wall? AT&T’s plan isn’t yet finalized, so we don’t know those answers. But it is difficult to see how such issues could be addressed without creating confusion for both consumers and developers.
There are many other problems with AT&T’s proposal, too. AT&T would surely have to bulk up its network operations to tackle the new challenge of monitoring which apps are being used when and how much data they are consuming. Those operations would also have to keep track of every app update to make sure they are monitoring the app accurately. And fraud would be a major concern: Just as some unethical content providers once disguised their goods to sneak them inside carriers’ walled gardens, underhanded developers could agree to AT&T’s terms but then tweak their apps to make their usage difficult to track.
Finally, there is the matter of AT&T’s policy of throttling data speeds for its most voracious consumers on unlimited plans. Enforcing that policy fairly would be a nightmare under the carrier’s proposed plan. App developers paying for that data surely would object if their apps were being throttled, but consumers who pay their own way would be outraged if their data transmissions were slowed while “premium” apps got a pass.
More bad than good
Admittedly, there are some attractive aspects to AT&T’s concept. Deep-pocketed developers and content providers might jump at the chance to pay for play, just as Fox News paved the way for success by paying cable companies $11 per subscriber to gain a foothold in its early days. Some consumers would happily embrace a model that granted them free data usage through specific apps. And developers like Netflix or Pandora that suddenly have skin in the game would have a new incentive to streamline their apps to use as little data as possible — a scenario that could benefit consumers, carriers and developers alike.
I think AT&T is likely to find interest from both developers and consumers if it brings its strategy to market. But the pay-for-play concept is simply too sophisticated and too costly to be effective in any substantial way. Instead, AT&T and its fellow carriers should continue to experiment with multiple solutions to handle the explosion in data usage, from throttling and data caps to strategies like congestion pricing (charging more for usage during peak times) and offloading technologies like Wi-Fi and femtocells. Narrowing the gap between data usage and data revenues will require the right combination of pricing models and network policies. AT&T’s proposed service might gain a little traction as an effective revenue stream for a handful of stand-alone, data-intensive apps, but it is highly unlikely to be a major part of any comprehensive solution.
Thanks for the comments, DS21.
I see your point that demand (based on use cases) could certainly impact consumers’ willingness to pay for access, but I think there are too many variables to actually build a billing model accordingly. For instance, I may not be willing to pay much to keep my kid occupied with Netflix on a Saturday morning, but perhaps I’d pony up in a big way for the same app but for a different use — when I’m in my hotel room with rotten (and expensive) Wi-Fi, for instance.
My point is that many (or maybe most) of us have varying demand for our apps based on fluid situations and use cases. So convincing us to pay monthly fees for preferred access through certain apps seems like a pretty tough sell.
That said, I can see where situational-based pricing for access might be a part of a larger solution. But I think that scenario differs from straight app-based pricing.
It’s tricky, but there is some validity to the argument. Most consumers believe that all data is created equal and we should be able to use it as we wish. Carriers would like to differentiate themselves and optimize their networks. Sure, putting up more towers and having “fatter mobile pipes” solves the problem, but that can’t happen overnight and even if it could it’s not always cost effective. Somehow there needs to be a compromise. Let’s say two people are using the same device on the same network and their apps are asking for similar bandwidths. Consumer A needs to log into Skype for an important video conference call. Customer B has handed his kid his device to watch Netflix. Undoubtedly Customer A would be willing to pay a bit more for preference on the network. If the network provider allocates the same bandwidth for both, Customer A will be upset. If Customer A is given preference, chances are that Customer B won’t care. Or, just maybe Customer B does not mind paying more so Sponge Bob is clearer for his kid.
But how is the carrier to know? Most consumers will likely be overwhelmed if we are presented with list of apps and are asked to allocate their bandwidth preference based on network congestion, time of day, usage etc. Tiered pricing needs to happen. One of the angles that might work is to have the app provider pay for the bandwidth preference. Netflix (Free App) or Netflix Plus (Pay $5 a month) so you get network preference.