This Startup Wants to Make Your Data Plan
Dirt Cheap
DataMi wants to let companies sponsor slices of your mobile data. But will it also break the Internet?
--
A few months back, I suggested to the rep at my mobile phone company that I was thinking about maybe, just maybe, giving up the $30-a-month all-you-can-eat data plan I’d had since back in grad school. That plan gives me the freedom to gobble up all the data I’d like using my iPhone. It’s a deal so delicious the company doesn’t offer it to new customers anymore. “Oh no, you hold on to that,” I remember the rep counseling, “You just don’t let a plan like that go.”
Hold on to it I did. But that rep and I were both outliers. The fact is that unlimited data plans are rather quickly becoming a thing of history. Companies like AT&T and Verizon are nudging users away from them — often as part of an upgrade to the latest smartphone or, as in my case, a switch to the latest innovation in family plans—in a bid to manage congestion on their sometimes creaky networks as we get into the habit of doing everything on our mobile devices. And so, we’re staring at a future of capped data and overage fees.
Into that bleak reality steps DataMi, a start-up based in the Massachusetts city of Chelmsford, just north of Boston. The company’s name is pronounced “data me,” as in my data, and it launched just a year and a half ago with the tagline, “Welcome to Free Data.” DataMi wants to give app developers, advertisers and content-makers the ability to pick up the data tab for mobile users visiting their websites or using their app. That might be one solution to our data-cap problem. The downside? Well, say some, it might just also be the end of the idea that on the Internet everyone gets an even shot.
Let’s set the stage. At the moment, cell phone users lacking an unlimited data plan pay a set price for a certain amount of data, say $40 a month for 3 GB of data. Every website you visit, video you watch, or email you send while using your cellular connection draws down from that data allowance.
What DataMi does is flip that model on its head. Instead of sucking 100 MB from my data bank to ‘pay for’ that Mary J. Blige album I bought, my music app could say, in effect, “This one’s on me.” It’s a bit like going to a bar with a budget for two, and only two, cocktails — but knowing that there’s a bourbon company rep at the bar willing to buy me a third. I’m happy. The bar’s happy. The bourbon company’s happy: if I can remember the transaction in the morning, they may well have a new, loyal customer.
In much of the world, that concept is called “zero rating,” a riff on an economics term for goods sold tax-free. In the United States, the phrase “sponsored data” is taking hold. But Mung Chiang, the Princeton University professor of electrical engineering who founded DataMi, would prefer we call it something else: “open toll-free data,” piggybacking on a concept familiar to anyone who has used a telephone or heard an infomercial in the U.S. in the last 50 years. Says Chiang, “People love 1–800-numbers.”
The fact is, the 1–800 Internet isn’t unheard of, either. Nearly a decade ago Amazon rolled out Whispernet, which covered the data demands of downloading e-books through the one-time cost of buying a Kindle. Since 2010 in El Salvador, Bangladesh, Zimbabwe, and scores of other emerging countries, Facebook has offered Facebook Zero, a data-free version of its site, with a pop-up touting that they can “Keep up with friends, share updates and send messages on Facebook with no data charges.” And just this June, the phone company T-Mobile began offering customers the option to “stop burning data” when listening to certain apps, including Pandora, SiriusXM and Spotify under a program called Music Freedom.
But what DataMi does that’s groundbreaking is to take that 1–800 power available only to the T-Mobiles, Facebooks and Amazons of the world — companies both savvy enough to negotiate with mobile networks and smart enough to make the technology work—and put it in the hands of app developers and website publishers of nearly any size. DataMi has negotiated deals with mobile network providers to make it possible for companies of all sizes to easily cover data tabs.
“The telcos weren’t designed to create these type of solutions in the networks,” says Harjot Saluja, a mobile industry veteran who became DataMi’s CEO in August of 2013. “We’re making it so easy that anyone can, in a matter of minutes, go on a portal and create a sponsorship for just two days of Thanksgiving.”
Soon various chunks of the data we use could be sponsored by all sorts of different advertisers, potentially making our mobile phone bills look like a NASCAR race car. That might not be pretty. But what would be far prettier is the price at the end of our bill. It could be just a fraction of what it is today: a couple bucks for the few bits of data no one has offered to cover.
If DataMi succeeds in the democratization of sponsored data, it won’t be without its critics. Some see sponsored data not as a clever solution but a cop-out that threatens the idea that to compete on the Internet you shouldn’t be forced to pay a gatekeeper. Are you going to keep using an app that drains your data budget when a similar one doesn’t? Raegan MacDonald, the European policy manager for the digital rights advocacy group Access, argues that “the real solution would be to provide data at a more affordable rate. But that’s not as fast or as easy.” There are echoes here of the net neutrality debate raging in the U.S. But if you believe that sponsored data violates the idea that all information on the Internet should be treated equally, that principle might well come with a cost to your wallet.
DataMi got its start as an academic research project at Princeton University’s EDGE Lab, headed by Chiang, a networks expert who is described admiringly by former Federal Communications Commission chairman Reed Hundt as “a volcano of ideas.” The insight that led to DataMi was that although cellphone use is placing considerable demands on mobile networks, that demand is unevenly distributed. Mobile operators could offer discounts for people willing to use data during non-surge periods, the way some electric companies offer cheaper rates to customers who forgo running their air conditioners on a handful of the hottest summer days.
In one early Princeton experiment, research subjects were given an iPad and told that it came with a data plan where each gigabyte cost 10 dollars. But during certain hours in the coming days, data would be significantly cheaper, up to 30% off. The users were given tools to make their data use more efficient: a stoplight-inspired dashboard showing hour-by-hour rates. They also had the ability to block Netflix during peak hours and to set a “delay tolerance” so that, say, a song they’d purchased at noon might wait to download itself at midnight.
The findings? Customers were indeed sensitive to time-dependent prices on data. Said one study participant, “[I] was less likely to goof off” when the iPad revealed that data costs were particularly high.
By September of 2013, DataMi had spun-off as a standalone company. Saluja, the new CEO, who had spent a decade in the mobile industry selling network capacity to mobile providers, was brought on board. The company has since grown to about 25 employees in Chelmsford, South Korea, Bangalore, and elsewhere. And it has raised $6 million in funding from venture capitalists and assembled a sparkling board of giants in media and in telecommunications. Alongside Hundt, appointed by President Clinton to head the country’s most important telecom agency, sits Marcus Brachuli, the former executive editor of the Washington Post and managing editor of the Wall Street Journal.
What’s so appealing? Two things, mainly, and the first is a technique. Whereas electric companies generally divide surge and non-surge pricing into hour or half-hour chunks, DataMi’s researchers figured out that it could find and make use of gaps when the networks weren’t being used to full capacity that were as brief as a half-minute. The company calls those moments “micro-nights,” no matter when they’re found in a day.
The second thing is a realization. As more companies follow the lead of T-Mobile, Facebook and Amazon and offer 1–800 versions of their services, there will, as Saluja explains, “be a huge incentive for app developers to leverage the concept of underutilized cells.” A translation: for me, an average customer, to shift my data use by 30 seconds might save a few bucks a month. For an audiobook company that streams its recordings data-free, pre-fetching chapters when data is cheap and slowing downloads at other times could save it massive amounts when applied to its whole user base.
But a key discovery gave DataMi added momentum. Users simply eat up more mobile data — and buy more songs, click through more news articles and play more video games — when they get some of their data for free.
Page Views
Last summer, DataMi conducted a study in partnership with Trove, a fledgling news curation site that spun out of the Washington Post. (Trove is today a project of Graham Holdings, a family-owned company that was once the Washington Post Company and is an investor in DataMi.) The study suggested that data caps are, as Trove CEO Vijay Ravindran puts, “a barrier to engagement,” as the amount that users interacted with Trove was closely tied to whether they were on WiFi or their cellular plan. By sponsoring data, Trove and DataMi were able to mimic WiFi usage rates for users who would have otherwise been chewing through their data allowances.
DataMi also tested out the concept abroad, in India, where cell phones are enormously popular but data plans far less so. Non-data customers were given access to some desirable piece of content, say the India vs. England cricket match telecast. For the sports network that aired it, the exposure was a boon, in part by enlarging the audience it could offer advertisers. For the network, there proved to be a “stickiness,” says Saluja. “If I use my mobile phone to watch a cricket match and once I did, I’m like, oh, this is kinda cool. I want to watch the hockey game [next].” They’ll buy their first data plan so that they can.
We’re in the very early days of figuring out the possibilities of sponsored data. For one of our phone chats, Saluja called while on his way to Costa Rica, where brick-and-mortar banks are toying with the notion of eating data costs for users of their mobile apps. “It’s so much cheaper to pay for someone’s data than it is to pay for them walking into a branch,” says Saluja.
Customers get phone-bill relief. Content creators get new users, and advertisers get more attention. Phone companies get a reprieve from the demand on their networks. There are those, though, who file this under “too good to be true.” Critics argue that sponsored data violates the principle that no one’s Internet traffic should be treated any better than anyone else’s. Otherwise the Internet’s winners and losers won’t be determined by merit but by the size of bank accounts. Indeed, such concerns have led Canadian authorities to tell mobile providers to stop offering their own video services data-free.
Sponsored data has also earned itself some high-profile critics; Bill Gates, for one, has dismissed it as “just a weird form of price discrimination.” To Raegan MacDonald, of Access, the question is how the public envisions the future of the Internet. A la carte mobile services might seem appealing, but, she asks, “in 10 years, do we want our Internet to look like cable TV?”
Saluja thinks such critiques are seriously off-base and ignore a key fact: that much of the commercial world runs on the same sort of framework that sponsored data brings to the mobile Internet. “If the argument is, how can Pizza Hut afford to spend more money on mobile data than Joey’s Pizza does,” says Saluja, “then we should shut down all kinds of advertising.” Other advocates for the model argue that imposing some sort of purity test on mobile phone data ignores the fact that the people paying for those ideals are the ones who rely on mobile Internet the most — the poorer among us, those in the developing world, school children. And then there is Reed Hundt, the former FCC chairman, who argues that customers should be aware that making use of every nook and cranny of the network has the same effect as stuffing a body into every airplane seat: It might be annoying, but it lowers customers’ costs.
At this point, that debate is largely academic. But Saluja thinks that it will inevitably become a practical question. The mobile Internet is expanding nearly every day; it’s not just phones or tablets—the smart car is, more or less, a mobile broadband device. “You’re not going to want to pay for a data plan for everything that comes into your life,” says Saluja. We might, say, opt to let the local tourism board cover our data costs when we’re in its city. The question is, which does the public think has the higher price: the data, or the changes to the Internet that sponsored data might bring along with it.