WSJ: AT&T Plan Would Let App Makers Pay for Subscribers' Data Use | Hacker News Reader

WSJ: AT&T Plan Would Let App Makers Pay for Subscribers' Data Use

one37.net

31 pointsONE3728 comments

I see too many problems with this idea.

(1) App makers that transmit data already pay for their web services' bandwidth. So they are basically getting double charged for the same data. (2) If app makers transmit data to other open services (Twitter API, Facebook API, YouTube API, etc.) then does the app maker pay for the subscriber's data use or does the API provider? Looks like a lot of room for confusion. (3) It shields the true cost of the subscriber's usage in much the same way I see company paid health care shielding employees from the true cost of health care. I would imagine this would result in AT&T raising their fees more and more against the app makers without subscriber's being aware and taking the blame on any increase in the cost of the app on the app maker rather than AT&T.

Given metered pricing, it's not really double-charging. Today the server pays for the packets to travel half way and the customer pays for them to travel the rest of the way. If the app pays for both halves then (in theory) the total charge is the same; it's just being paid by a different party. I doubt that AT&T would charge an app the same ridiculous overage fees they charge customers, though.

Regardless, the market distortion caused by the cost-shifting is real and problematic.

> If the app pays for both halves then (in theory) the total charge is the same

It sounds like you're assuming the carrier (AT&T) would stop charging the consumer for the bandwidth as well. As in AT&T would offer a data plan without a data fee. I'd wage that AT&T's gambit is to get paid twice for the same carriage, and not what you have in mind.

"A feature that we're hoping to have out sometime next year is the equivalent of 800 numbers that would say, if you take this app, this app will come without any network usage."

So, in other words, unless you're already rich enough to pay for your customers' data usage, your app will be at a disadvantage on AT&T's network. This is why we need source-neutrality in networks. Otherwise existing players can subsidize their services and "dump" their product in order to drive new competitors out of the market.

> This is why we need source-neutrality in networks.

No, we do not. There are already laws on the books against "dumping" their product to drive out competitors. It's called price bashing and it is illegal.

If Netflix is able to subsidize their data usage and remain profitable, great. If, on the other hand, Netflix decides to go with this to kill off the competition with the calculus that they will lose money temporarily and hike prices after the competition is gone ... well, that's illegal and on top of everything, won't even work.

Also, AT&T's ploy here won't work either. They are trying to increase data usage while hiding the cost of that usage in 3rd party fees. Yeah. Watch content providers offer two price tiers. One for AT&T and one for non-AT&T. Won't work and a waste of time. You're just shifting costs.

Also, 800-numbers became popular during a time when every street corner had a pay phone. It was about actually letting the customer contact you from anywhere, not about shifting costs.

There is no need for regulation. The market will take care of this stupidity by itself.

Sounds great on paper. In real life there's a certain degree of collusion between the major carriers, and it's possible AT&T could convince Sprint and Verizon to play the same game. It's also possible one of them had the idea to begin with, and AT&T won't even be the first to implement it.

After a couple of years you might get a class action lawsuit far enough along that the FCC says the whole thing is perfectly legal and it doesn't see the problem. Or the carriers might backstab each other for secret reasons. But this matter cannot be trusted to the market.

I believe it's actually called predatory pricing & is covered under antitrust. But it's not that easy to prove. It is also a regulatory practice to keep the market competitive.

Additionally offering two tiers may seem easy but you risk pissing off a potentially large portion of your customer base & the ISP who's name your sullying in the public space. It takes some balls to go out and say "AT&T are being dicks so you have pay more", some companies will not do that & would rather do some backroom deal that negotiates a lower rate than the sticker price AT&T is asking. The result is still that AT&T get's more money in the end.

This is the equivalent of free shipping. I predicted it a while ago (patting back): http://richvsreach.wordpress.com/2008/06/03/the-long-game-on...

What it’s not is net-neutral. I’m OK with that.

Good call. I called it too, much later than you though.

http://blog.altudov.com/2011/08/27/the-upcoming-amazons-andr...

The cost for data does not dissipate, it is merely redistributed. Rather than giving your money to AT&T for your service, it will just be charged by data-heavy services - particularly as this would increase their data costs.

Furthermore, reliance upon such a system would likely give AT&T license to reduce data package size, thus harming the average smartphone user.

Although AT&T might frame this as innovation, such maneuvering is transparently self-serving. Rather than having the customer blame them for high costs and decreasing data allowances, AT&T can place the blame elsewhere.

Knowing AT&T as we do, the potential for good in this model is rather woefully undercut by its obvious potential for bad.

So instead of giving up on the flawed idea of data caps, they are putting the onus of caps on app makers instead of customers? I can only imagine that this new spin on an old strategy will have all the same problems.

Perhaps this technical post from AT&T labs (http://bit.ly/zFyXTw) discussing their work with Pandora to optimize power & bandwidth consumption might shed some light on AT&T's thinking. It might be a win-win for AT&T to partner with such app producers, optimize their network utilization in exchange for some end user subsidy -- extending the subsidy carriers give to device manufactures to the app developers, essentially?

This is Amazon's Whispernet tied to service instead of device, and it is certainly, "fair."

We don't expect users to pay service fees to get Kindle books delivered.

I think this is really a smart move, and would create a brand new revenue stream for the carriers coupled with (ideally) cost-savings for the customer. It'd also encourage efficiencies by app developers in how they use the network.

It would be REALLY interesting to flip this model.

In other words, AT&T redistributes a portion of their subscription revenue to app makers based on the bandwidth they get people to use.

Then, app makers would have a direct financial incentive to drive usage, without all the hassles of explicit billing.

But more bandwidth usage is exactly what AT&T doesn't want.

What a messed up system. AT&T is in the business of selling bits. Most sellers want to sell more of what they sell, not less...

No, AT&T is in the business of selling _access_. The less each customer costs them, the better.

The problem is when you sell X bits for $Y, you'll want to keep X as low as possible and Y as high. Now if bits were fully metered they'd have no problem selling you all they have.

No, AT&T is in the business of selling phone contracts. The actual bits themselves are loss leaders to help sell those subscriptions.

OK, you're right. Maybe it was wishful thinking: they should be in the business of selling bits.

That would encourage people to make their apps bandwidth inefficient

Should restaurants pay for the gas I use to get there?

This is a poor analogy, because you don't pay a flat monthly fee for your gas usage. The customer is already incentivized not to drive too far to eat at a restaurant because of the fuel expense and the time required to get there.

There are few analogies that work well outside of examples involving utilities, but most utilities are metered rate. The carriers built their current pricing models based on a usage profile that is rapidly disappearing. Basically, they priced their service for a model where the user rarely uses the actual service. Now they're desperately trying to escape their own trap.

Carriers believe that users don't want metered service, and in fact, they don't. Why would I want to expose myself to additional risk when a flat rate service is already available? Carriers captured massive profits when cellular users didn't utilize their plans, but aren't prepared to take the hit when consumers actually use what was sold to them. It's a sham.

That actually doesn't seem like such a bad idea. Well except that the connections should have been unmetered like residential broadband is (in most of the world) in the first place...

Can someone summarize the actual WSJ article or submit a non paywall link?

What the hell? I already pay for my data. AT&T is just making a blatant money grab.

They need some competition.

Reply on news.ycombinator.com