Despite Limited Interest In AT&T's Sponsored Data, Company Still 'Bullish' On Its Awful Precedent
from the creative-gatekeeping dept
After hinting about such a project for some time, you might recall that AT&T introduced "Sponsored Data" at the company's developer summit around this time last year. It works like this: if companies pay AT&T a fee their content is specifically allowed to bypass AT&T's already entirely arbitrary (as in, not tied to any real world costs or network conditions) usage caps. To hear AT&T pitch it at the time, this would be akin to "free shipping" or a 1-800 number for data, and an incredible boon for consumers who want to conserve their pricey data allotments.While some consumers seemed quick to applaud the idea, they weren't understanding the awful precedent AT&T was setting. If you allow AT&T to set arbitrary caps then charge companies to bypass them, you're injecting a company with a rich history of anti-competitive behavior into a content and service ecosystem that works much better with it out of the way. Also, as VC Fred Wilson correctly noted at the time, such a model puts smaller companies and developers at a distinct disadvantage to their deeper-pocketed counterparts. What AT&T pitches as a great creative boon to industry is actually AT&T just desperately trying to retain gatekeeper power.
While AT&T executives have spent two years claiming that interest in this idea is through the roof -- one year later, just ten (mostly smaller) companies have signed up for AT&T's pilot. While Sponsored Data played a starring role at last year's AT&T Developer Summit, executives didn't mention the project once during this year's event. To hear AT&T tell it, there's still tremendous interest in the idea -- despite the fact there's clearly not tremendous interest in the idea:
"Nonetheless, AT&T CMO David Christopher told FierceWireless that the carrier is still "very bullish" on the program...What we said last year, and what we've continued to say, is Sponsored Data is a really unique, interesting capability that is going to take time for it to evolve into various business models," Christopher said in an interview. "We are seeing interest from a variety of developers and content owners in Sponsored Data."While some companies aren't eager to court net neutrality controversy, others seem entirely oblivious to the threat such a model poses to innovation and smaller developers. Beyond just the obvious neutrality implications, the idea doesn't appear to be gaining traction with companies because new wireless shared data plans have most people signing up for significantly much more data than they need in order to avoid costly overages. In other words, when you have more cellular data than you need, and you're spending a lot of additional time using Wi-Fi, having a few apps or ads that don't impact your data allotment doesn't mean all that much in practice.
As such, it seems like only a matter of time before AT&T mutates the Sponsored Data idea into something notably more awful with a better sales pitch. As I've noted previously, while most of the net neutrality discussion focuses on outright blocking of websites or throttling of connections, the real danger zone is these kinds of "creative" pricing efforts where carriers try to use their gatekeeper power to desperately avoid being dumb pipe providers. It's here, under a glossy coat of PR paint where the real neutrality violations are going to occur, but as we've seen, it's difficult to craft neutrality rules that protect consumers from obnoxious shenanigans -- while allowing for real pricing and service experimentation (should that actually happen in the broadband sector someday).
In this case, we appear to be just lucky in that AT&T's implementation was just so bad most companies were bright enough to steer clear. That's not always going to be the case. As we've seen with the positive reaction to T-Mobile's decision to let the biggest music streaming services bypass its cap (which of course hinders smaller companies or nonprofits not big enough to get whitelisted), it's very clear it's possible to create new business models that tilt the playing field and screw smaller companies and consumers -- all while receiving thunderous applause for the effort.
Thank you for reading this Techdirt post. With so many things competing for everyone’s attention these days, we really appreciate you giving us your time. We work hard every day to put quality content out there for our community.
Techdirt is one of the few remaining truly independent media outlets. We do not have a giant corporation behind us, and we rely heavily on our community to support us, in an age when advertisers are increasingly uninterested in sponsoring small, independent sites — especially a site like ours that is unwilling to pull punches in its reporting and analysis.
While other websites have resorted to paywalls, registration requirements, and increasingly annoying/intrusive advertising, we have always kept Techdirt open and available to anyone. But in order to continue doing so, we need your support. We offer a variety of ways for our readers to support us, from direct donations to special subscriptions and cool merchandise — and every little bit helps. Thank you.
–The Techdirt Team
Filed Under: net neutrality, playing favorites, sponsored data, zero rating
Companies: at&t
Reader Comments
Subscribe: RSS
View by: Time | Thread
Sponsored Data IS Throttling
If someone can pay to get better routing of their packets, then someone else's packets are being throttled unfairly.
If not, then there is sufficient network capacity and no Sponsored Data would be necessary.
So which is it? Is AT&T not building sufficient network capacity to deliver what its own customers pay for (eg, Netflix to my home)? Or is AT&T going to unfairly throttle other internet services to favor those (like Netflix) who might pay AT&T to buy what AT&T's customers already are paying for: getting netflix to my living room.
AT&T, here is a free clue: Build out your netnwork capacity to deliver what your customers are paying for. Charge your customers enough to deliver what they want, and to make a profit. Isn't that simple? That almost seems like how all businesses should operate.
[ link to this | view in chronology ]
Re: Sponsored Data IS Throttling
It would naturally happen with competition.
[ link to this | view in chronology ]
Re: Sponsored Data IS Throttling
[ link to this | view in chronology ]
Creative routing?
[ link to this | view in chronology ]
Lets see how bullish they are...
[ link to this | view in chronology ]
[ link to this | view in chronology ]
Re:
The RIAA would also be happy, because at least none of the money went to the artists.
[ link to this | view in chronology ]
Domino Theory Isn't Right
It seems to me that all the good arguments against sponsored data rely on the Domino Theory. That is, "Well, this is barely OK, but it's just one step away from being wrong. Therefore this is as good as wrong." But, much as I dislike sponsored data, I can't agree in Domino Theory criticisms. It's like thought crime, and you're the precog:
"it seems like only a matter of time before AT&T mutates the Sponsored Data idea into something notably more awful"
Now, you argue that, based on their history, AT&T can be expected to take the next step, and topple the next domino, entering into shady bad-guy territory. I agree with this. But...precog argument again...I don't believe in pre-crime policing.
"some consumers...weren't understanding the awful precedent AT&T was setting" - but precedent to crime isn't a crime.
I think freedom dictates that we let the stakeholders to the borderline shady plays. Now, it would be better if there were more competition, so we could vote with our wallets, but that's another problem on which we agree.
"[the sponsored model] model puts smaller companies and developers at a distinct disadvantage to their deeper-pocketed counterparts."
Perhaps. But what about a startup that has a service targeting a segment of consumers that just don't have data plans, or have tiny ones. Let's say I'm a startup with one service/app aimed at seniors, and another aimed at latino immigrants. I know my target market has particularly low subscription rates to mobile data. With sponsored data, I have the flexibility to create a biz model where I pay the freight, and can actually have a market to address. In this case, Sponsored Data enables the biz model.
The important thing is that the network is still neutral, meaning that the company that pays for Sponsored Data doesn't have their packets prioritized over a competitor that chooses a more conventional model. This is unrelated to whether the user pays the freight or the content company. I know, you don't trust AT&T not to prioritize in some sneaky way...and you may be right. But then your beef is against the prioritization, not the payment model.
I'm glad this idea is not taking off. I hope it flops. But I think it's a fair idea that opens up different biz models, and that's OK.
In short: For fair biz, each packet needs to be billed the same fair rate, somebody needs to pay, and Sponsored Data should get the same prioritization as regular data. So long as somebody pays, there is no preference, there are just different biz models.
[ link to this | view in chronology ]
Re: Domino Theory Isn't Right
But this is exactly why it's a scam. The entire thing is a scam. There is absolutely no reason to pay for "packets." It doesn't matter how much total data I use.
Here's an example. Let's say the capacity of AT&T's network is some random number, like 100 gigabytes per second (purposefully not using bits). Let's say all other users with me on that network are using 70 gigabytes/sec. I use (in my imaginary world with impossible speeds) 1 gigabytes/sec of that, maximum.
If I use all of my bandwidth all the time, AT&T is still under capacity. All it costs them is the tiny fraction of my bill that pays for the electricity needed for my portion of data transfer. If I download at max capacity every day for a month, I download roughly 2,592 terabytes of data. If I download at max capacity for a week, and then shut off my computer for the rest of the month, I only use about 60 terabytes.
Other than the difference in electricity, which again is miniscule (and more than covered by your bill, either way), AT&T isn't paying a cent for the 2,500 extra terabytes you downloaded. The network pays for it's bandwidth capacity (the maximum speed it can sustain at a given point in time) not it's network throughput (how much data actually passes through the network).
This is the lie that makes the ludicrous seem reasonable. If you accept the lie that having, say, Netflix pay for your 250 gb of movie watching means you don't have to pay for it that sounds good for you. If you accept the lie that Pandora is now paying for your streaming music, it sounds good for you.
Unfortunately, the lie is the key. Without the lie, important questions start getting asked, like, "well, if I can stream Pandora at max bandwidth all the time without overloading your network, why can't you stream, say, Sony Music instead?" Or "if I'm paying for my 50 mbps download rates, why does Netflix also have to pay for the capacity I'm already paying for?" And finally the most important question: "if using more data than I'm allocated costs more money, why doesn't using less mean I pay less?"
Right, because it's a BS metric. Now, before techies start arguing about how this isn't 100% true (transmission costs, maintenance, technicians, and improving infrastructure all have non-zero costs) keep in mind that most of these costs are independent of the amount of data transmitted. You'll see analysis indicating that transmitting 1 gb of data costs anywhere from 1 cent to 10 cents but most of these calculations are based on average bandwidth infrastructure and the numerous fixed costs above.
The key point is that those costs exist whether an individual uses 1 gb of data or 1 mb of data, making the metric mostly imaginary (sort of like calculating the miles per gallon efficiency of an electric car; while it may give you a perspective on efficiency, it's ultimately comparing apples to oranges and doesn't accurately reflect the comparison between the two).
"Fair biz" only works when the game isn't already completely rigged in favor of the house.
[ link to this | view in chronology ]
Re: Re: Domino Theory Isn't Right
But there is a very real investment in the network infrastructure, and also an ongoing expense in operations. And when existing network capacity is filled up, significant capital must be invested to increase capacity for the newer traffic.
And there's nothing theoretical about that. Data traffic has grown phenomenally year over year, every year. We could not carry all those bits we send today using 2008's network assets, nor 2012's.
AT&T spent $19 Billion in one year on their network. That's real money. Separately, how do you think companies like Cisco, Ericsson, HP, Huawei, Alcatel, etc. all make their big revenues every year? That money comes from YOU, from your per GB fees, through the carrier, and on to their equipment suppliers. Ericsson will not provide network upgrades for free, so AT&T has to charge those who use data to pay for the upgrades.
So, a fair way to pay for a GB of data is based on the average cost (in operational cost and capital rent) of that GB of data, plus some profit margin.
Sure, the US carriers margins are too high, because competition is too light. But the basic notion that a GB has a cost and a price is entirely fair, and not just some wacky imaginary construct to rape the customer. It's not the only way to charge for the network, but it's an OK one.
If you'd rather pay some rent for some network capacity, then that's just another model. It might work. You'd end up paying near the same amount, though. It would be harder to allocate the higher payments to the people that use more, too.
I would say the construct that "transporting a packet is free" is the fictional construct. It's only true if significant investments were made, and continue to be made, in over-capacity of infrastructure. You make a lot of assumptions there, and you're making them with someone else's capital.
"The network pays for it's bandwidth capacity not it's network throughput."
True. But...
Analogy time:
Restaurants have very high fixed costs, and some marginal costs in labor and food.
Do you walk into a Subway and say "I won't give you $6 for this sandwich, that's a farce. The Sandwich is $3 fixed costs, and $3 variable costs. The fixed costs are already done and paid for. I should only pay you the $3 marginal cost for my food and labor inputs."
Cuz that's what you're doing to ISPs.
Listen, ISPs oligopolists are dicks. They're not easy to defend. But occasionally they're right.
[ link to this | view in chronology ]
Re: Re: Re: Domino Theory Isn't Right
It's more like paying $6 for the sandwich, then Subway charges you $3 more if you want it in the next hour because there's a long line. Then they hand you half the sandwich because they didn't stock enough ingredients and other people already used them. Unless you ordered a ham sandwich, because the company that sells ham paid extra, and now they suddenly have enough ingredients for a whole sandwich.
The issue is that with the ISP is that it's pretty much all fixed costs. If the ISP is going to build an infrastructure to support a minimum level of service, the money they pay into it is effectively a fixed cost. They can't pay the technicians less, or buy less of their equipment, or install less lines.
I'm not saying that the fixed costs don't exist. I'm saying that the fixed costs are reliant on meeting the demand for bandwidth capacity at the maximum capacity required. If less bandwidth is used, the ISPs still have to pay the same fixed costs.
That's why paying a monthly fee for a certain bandwidth capacity makes perfect sense. If I want 50 mbps, I pay the company for that capability, and I'm paying to account for their fixed costs. They now have a fixed revenue stream to account for my requirement.
So what you're telling me is that ISPs are operating at a loss of fixed costs, and having to charge extra for imaginary costs to make up for it. I pay $110 per month for 100 mbps internet access. I rarely get 100 mbps, but I usually stay over 50. You're saying that my $110 per month, along with all the other customers, is not enough to cover their fixed costs?
AT&T spent $19 Billion in one year on their network.
Yeah, so? AT&T also received over $128 billion in revenue, meaning less than 15% of their net income went to improving their network. And you're telling me they can't afford to spend more on giving me the service they've advertised?
Maybe if I were spending $20 per month on my internet and they needed extra revenue to cover the network upgrades and still make a profit I might accept their logic, even knowing the charges were bogus. But since I know they have a huge income and less than a fifth is going to improving my service, I'm not paying a fifth of the price, and there's no technical reason to charge per gigabyte when I'm already paying for bandwidth in megabits per second, I can't wrap my head in the knots required to eat their bull**** sandwich.
Sorry.
[ link to this | view in chronology ]
Re: Re: Re: Re: Domino Theory Isn't Right
No. I clearly said they make high profits, because of a lack of competition. But we're not discussing what fair profits are, we're discussing pricing models.
" If I want 50 mbps, I pay the company for that capability, and I'm paying to account for their fixed costs"
Yeah, but you probably know that they oversubscribe their customers, and estimate actual usage for network planning purposes. You are not paying for 50 Mbps. You are paying for "up to 50 Mbps, best effort". You do not have dedicated capacity, as business lines often do. If you want 50 Mbps with an SLA (Service level Agreement) to guarantee and dedicate that full capacity to you, then expect to pay hundreds of dollars per month.
Network planning has always been thus (oversubscribe), because it would be stupid to build a network any other way. Roads, toilets at the theater, elementary classrooms, theme park rides, gym memberships. They're all built assuming everyone doesn't use it at once.
[ link to this | view in chronology ]
Re: Re: Re: Re: Re: Domino Theory Isn't Right
Now, if the gym were an ISP, they'd also say I can only spend 20 hours per month in the gym, and this is fine, because the average member spends less than 20 hours per month. It's only those few powerlifters that spend longer, and they get charged $10 extra for every 5 hours they spend in the gym extra. Also, once they reach their limit they have to wear a lanyard that gives priority to all other gym members because they've already used their share.
And that's the current situation. ISPs are proposing some "innovative" solutions. Now, Cybex Fitness Equipment can pay the gym a monthly fee and so any time you spend on their equipment doesn't count towards your monthly hour limit. Also, all equipment companies need to pay a monthly fee to the gym for their floor space, because they shouldn't get to use the gym's infrastructure for free. And since Precor paid some extra money, their equipment can be used as long as you want; everyone else has a one-minute maximum use time.
But hey, there's plenty of competition for this gym. In fact, there are four other gyms in town! They all only have two treadmills and single weight bench, but hey, you have options for your workout needs! Some other big gyms are available in other cities, but due to contracts with those gyms, only one of the big gyms exist in your city. Coincidentally, those gyms give their customers the same service and pricing.
When I'm paying for a possible maximum capacity with a monthly fee I'm sort of bothered by a bunch of random fees and restrictions that aren't actually caused by any technical restriction. The gym has a limit to its capacity, and during peak hours you'll probably have to wait to finish your workout. But that limit exists regardless if you spend all your hours in one energy-drink fueled 20-hour marathon or 20 days of one hour workouts. And the exercise equipment is the entire reason you want to buy a gym membership in the first place.
I'm not saying that ISPs should be required to give me my advertised speed at all times. Nobody is (or should be) saying that. I just want my limit to be caused by the capacity of the network, NOT some arbitrary data limit that has little to no impact on my speed. And the whole reason I go online is to access websites, so I don't understand why the websites have to pay my ISP extra so I can access them. I'm already paying for that (and they're paying their own ISPs to be there).
If they weren't making insane profits, I could sort of understand it as a method to make up the money they're losing because actually charging me for what they're selling isn't enough to cover their costs. But that's clearly not the case.
Since competition doesn't exist, and because they're trying to charge me and everyone else for imaginary resources, I believe they are engaged in anti-consumer, monopolistic, behavior that is not required for the health of the network or for their own business requirements and therefore should be regulated. And unlike a gym membership, for the majority of Americans internet access is not optional. We use it for business, we use it for networking (the social kind), we use it for knowledge, and much more. Living without internet access in the modern age is almost akin to living without electricity; possible, but miserable, and suicidal from a business standpoint.
We need the internet to stay accessible to all people, and to prevent the fact that access is necessary to allow monopolistic business to abuse their customers. It's sort of like banks in many ways, and I doubt many people would argue at this point that banks should be unregulated.
[ link to this | view in chronology ]
Re: Re: Re: Re: Re: Re: Domino Theory Isn't Right
Yeah, that's the crux of the matter right there. If there were real competition in the sector this sort of activity would be fine. Arbitrary limits on usage can reduce network saturation. The result for the end consumer is a less useful and more expensive service, but that would be fine, if there were alternatives.
However, ISPs (both wired and wireless) have spent the last two decades or more deeply entrenching themselves. The networks they operate were often subsidized, at Federal, State, and Local levels. They have spent millions of dollars lobbying (successfully) for anti-competitive laws of their own design. They have either natural or government (Local or State) granted monopolies in most of the regions they operate in. They collaborate with their so-called competitors, dividing territory and colluding on prices and practices. All while providing what few would disagree is a basic necessity of modern life.
When all the significant providers of a necessary service engage in collusion and anti-consumer behavior, it is, and rightfully should be, time to regulate that industry. I don't lightly suggest regulation. Careless or unnecessary regulation can have enormous costs and serious repercussions. However, ISPs have shown time and again that, like the banking industry, they will engage in anti-consumer behavior for so long as they are permitted to do so. It's time to tell them otherwise.
[ link to this | view in chronology ]