Last spring, Facebook founder Mark Zuckerberg invested in an impressive domain name: internet.org. Then, in August, he posted a video featuring snippets of John F. Kennedy’s “Strategy of Peace” speech and blogged that he would “share a rough proposal for how we can connect the next 5 billion people and a rough plan to work together as an industry to get there.” With that, Facebook and six corporate partners—including Nokia, Samsung, Qualcomm, and Ericsson—became part of a swelling movement of tech companies declaiming a commitment to connectivity, seemingly moved by the fact that only 2.7 billion of the world’s seven billion people have Internet access. In October, Google helped launch the Alliance for Affordable Internet (whose members include Facebook and Ericsson). It is pushing for cheaper Internet access through policy and regulatory reforms.

Behind the focus on the world’s unconnected lie some complicated realities. The companies involved tend to emphasize delivering more data to people who already have network access rather than extending communications connectivity to people who have none. And despite Zuckerberg’s lofty statements, Facebook in particular is falling short of some of Internet.org’s goals: the company isn’t investing in network extensions in developing countries, and its business practices, in many cases, have obligated Internet service providers in such places to incur extra costs.

Internet.org is still more of a press release than a plan. But its first formal statement, a 74-page white paper cosigned by base station maker Ericsson and chipset maker Qualcomm, is telling: it sets a goal of delivering data 100 times more efficiently to mobile phones, the devices most Internet newcomers will use to link to the Net.

Increasing efficiency is a perennial goal. And if it makes it possible for ISPs to offer broadband more cheaply, it could make people better off. (Research from the World Bank says that increasing broadband penetration in developing countries by 10 percent boosts their annual economic growth by 1.4 percentage points.) But getting people more data faster is quite a different objective from introducing connectivity in the first place.

Ground truths

Facebook is a major online presence around the world. Take Africa, where it often ranks first or second in popularity among websites. Yet Facebook doesn’t have data centers there, which means content generated by Facebook members in Kenya, for example, has to traverse undersea fiber-optic cables to data centers on other continents. That costs local ISPs at least $100 per month for each megabit of traffic. This charge wouldn’t apply if Facebook stored user content locally.

The ISPs pass those extra costs on to consumers—which surely can’t help Internet expansion efforts on a continent where only 16 percent of people have Internet access, compared with 39 percent worldwide. “It’s a bit disingenuous,” says PharesKariuki, who runs Angani, a cloud computing startup in Nairobi. “On the one hand, Facebook claims to want to give Africa access through Internet.org, but when it comes to the business decisions they are making, as far as Africans are concerned, I have not seen anything that reflects that value yet.” (It is worth noting, however, that Akamai, the Web optimization service, is establishing infrastructure in more and more African locations. To the extent that Facebook uses Akamai’s service, it reduces the extra costs that ISPs in those regions would incur.)

As part of Internet.org, Zuckerberg published a white paper titled “Is Connectivity a Human Right?” in which he wrote that the company has “invested more than $1 billion to connect people in the developing world over the past few years.” But the details were absent: spent on what, to connect whom, and to what? Through a spokesman, Zuckerberg turned down an interview request. But on closer inspection, that statement apparently means “connect people to Facebook.”

Facebook spokesman Derick Mains e-mailed a clarification: the company, he wrote, hasn’t invested in any “physical buildout of infrastructure” to connect people. He declined to say where the $1 billion went, giving only one example: Facebook’s $70 million purchase of Snaptu, whose technology makes it possible for apps like Facebook’s to run on the basic phones that are common in developing countries.

Such acquisitions, of course, are meant to improve Facebook’s own operations: the company, like others, is keenly interested in having its service accessible on as many phones as possible. Facebook is also doing important work to develop ways of delivering information more efficiently to smartphones that run the dominant Android operating system, says Jay Parikh, Facebook’s vice president for infrastructure.

Facebook will surely come up with technologies that are useful on all kinds of mobile phones. But Ethan Zuckerman, who has helped lead several Web projects in poor countries, says that “to wrap that into a press release that turns Facebook into the savior of the developing world is hard to swallow.”

Tapping the airwaves

Other Internet companies have gone much further, funding Internet infrastructure projects that also happen to advance their own interests in getting more people to use their services.

One is in the capital city of Kampala, Uganda, a metropolis where you can get relatively slow connectivity from any of about 10 mobile carriers or Internet service providers. In November, Google announced that it had installed 170 kilometers of fiber-optic lines in Kampala, a major step forward that could enable local carriers and ISPs to provide faster speeds at lower prices. (Fewer than 1 percent of sub-Saharan Africans have fixed broadband, defined by the U.N.’s International Telecommunication Union as a data rate of two megabits per second; 11 percent have mobile broadband, defined as 3G or similar service.)

A handful of other projects are meant to provide Internet access where none previously existed at all. One is unfolding in the region around Nanyuki, Kenya, a town at the foot of Mount Kenya. In poor and sparsely populated areas like this, extending fiber makes no sense economically—wireless carriers often fail to recoup their investments in even conventional cellular base stations powered by diesel generators. But in Nanyuki, an experimental low-cost wireless Internet system is radically altering the economics.

It works like this: first, a powerful microwave transmitter delivers a high-bandwidth connection from a fiber terminus to several fixed wireless base stations over tens of kilometers. These base stations retransmit data on unused television frequencies—called “white spaces”—to 40 solar-powered Wi-Fi routers and phone-recharging stations in schools, clinics, businesses, and community centers. The Nanyuki apparatus already serves 20,000 people, and this capacity is set to triple. Most important, it does so for less than $5 per user per month—5 percent of the region’s average annual income
of $1,200.

The company behind this effort is Microsoft, but Google has just completed a similar trial to provide bandwidth to schools in Cape Town, South Africa. Companies are testing many other white-space efforts around the world. The impact could be large: what many places need is simple access to the airwaves, which is frequently restricted by national governments. “If you look around the world—whether in the U.S. or the Philippines—the issues around digital inclusion and universal access are mainly policy challenges,” says Paul Garnett, director of Microsoft’s technology policy group.

Reaching the remotest

But white spaces are only as good as the base stations and power supplies at the farthest endpoints. Vanu Bose, CEO of a company called Vanu that develops cheap cellular base stations, tells the story of an enterprising man in Zambia who collects cell phones every morning from his fellow villagers. He then drives three hours to a spot where he can get a signal from a cellular tower—and switches on all the phones so they can ingest all the text messages and voice mails that have accumulated since the previous day’s excursion.

This workaround is a reminder that there are still more than 200 million people in Africa alone who don’t even have the most basic cellular phone service. For Zambia, Bose has developed what he claims is the lowest-power base station on the market: a rugged unit that can connect to the Internet in a number of ways, including microwave links, satellite links, and white spaces, and serve up access to 1,000 villagers per node. All it needs is 50 watts of power from solar panels, with a few watts left over for a communal phone-charging dock. This provides very basic voice and data service and maybe one low-bandwidth Wi-Fi hot spot.

Broadband it ain’t. But such service can be transformational—enabling families to stay in touch, emergency medical aid to be summoned, educational materials to be delivered. “Internet.org is all about higher-capacity networks and more bandwidth,” Bose says. “But we shouldn’t think about bandwidth first but connectivity of any kind first. They are very different things. One communications transaction per day is infinitely better than zero.”

Beyond hyperefficient setups like Bose’s, Google has prototyped a new concept: fleets of solar-powered balloons in the stratosphere, networking among one another and beaming Internet connectivity to far-flung rural areas at speeds comparable to 3G. It’s been criticized as a marketing stunt, and it may not even work. But in contrast to Facebook’s effort to increase data efficiency, “at least it’s funky and new, at least it’s interesting, at least it’s ambitious,” says Ethan Zuckerman, who today is director of the Center for Civic Media at MIT’s Media Lab.

Facebook says its focus is in the right place, and that helping more people who already own phones to afford data plans is a crucial job. That’s why the broad outlines of Internet.org involve figuring out how to deliver data more efficiently, in part through new business models. “A good way to look at it is that it’s a first step, and a really hard problem to solve,” says Aaron Bernstein, a former Qualcomm executive who is now a director of mobile partnerships at Facebook. And all the companies and organizations promoting and working toward Internet connectivity agree that there will be no silver bullet. “Only a lot of lead bullets,” as Facebook’s Parikh puts it.

But Facebook must shoot those bullets at the right targets. If the company really wants to make access more affordable, it can make sure its data is in the countries where people are using the service. If it really wants to connect more people, it can fund and support cutting-edge wireless networks. As John F. Kennedy said about the Peace Corps, 24 years before Zuckerberg was born: “Americans are willing to contribute. But the effort must be far greater than we have ever made in the past.”