Droidcon boston nathan eagle header

Why Android Will Get the Next Billion People Online

More than 4 billion of the roughly 7 billion people on Earth now have the ability to get online, thanks in large part to the rise of Android. Yet, there are 1.3 billion smartphone users living in emerging market countries like India or Indonesia who are limited by the relative high cost of mobile data. 500 MB of data costs a user in India 17 hours of minimum-wage work in India. In the United States, the same amount of data costs only 3.5 hours worth of minimum-wage-work. But, the hunger for Internet access and smartphones is not slowing down. By 2020, emerging markets will have more than 2.5 billion smartphone users and the demand for Internet access will increase by 500%.


I’m Nathan Eagle and I run a company here based in Boston called Jana and we are now the largest provider of free internet in emerging markets. This is a pretty awesome thing to say because we started the company back in 2009 and it was challenging. Basically, we’re trying to make a pitch that we were going to become a big profitable company by giving away a free service to some of the poorest regions in the world. For raising capital that turns out to be a challenging story to tell.

As we’ve gained momentum and as this rising tide has happened where you’ve started to see this proliferation of $20, and $30 Android handsets now in these emerging markets, the story becomes I think more plausible. I’m really proud that over last year, we gave over 30 million people free internet access in markets like Brazil, India, and Indonesia. On the fundraising side, we’ve now raised nearly 100 million dollars to push this Venture into what I think is a world-class company.

I was going to talk a little bit about what’s going on in these other markets. I think, it’s really easy, especially these types of conferences, to get into this navel-gazing mode about how awesome this new app or thing would be, and how it would improve my life. Whereas, it is important to realize that the vast majority of people using these Android devices, don’t live in North America or Northwestern Europe. The vast majority are in emerging markets, in countries like India, Africa, Kenya, Nigeria, and Brazil.

My Background With Mobile Carriers(2:10)

We started this venture by accident. I was at MIT at the time. My background is data analytics and specifically, I do data analytics for mobile carriers. I would analyze what’s called call detail records, or CDR. These are the types of data that have gotten into the press a lot over the last few years around at the NSA. Carriers have been analyzing these types of data sets for a very long time, well before the revelations that Snowden brought the table.

As an academic at the time, I would go in and help the operators build models of churn and proctored option data but the thing that they cared most about was ARPU, Average Revenue Per User. ARPU is plummeting in the emerging markets and it should make sense when you think about it. As these handsets are getting cheaper, essentially the average mobile phone subscriber’s getting poorer, because poorer people can suddenly access these handsets.

Carriers are really struggling to figure out what to do with the ARPU problem. They’ve got a tremendous amount of pressure under their boards to try to figure out how to invest in 4G and various other things. How do you raise the capital required to be competitive in a market where your consumer is getting poorer on average?

SMS Blood Bank(3:48)

I didn’t have an answer to that at the time, but I ended up taking a sabbatical from MIT going on to teaching mobile phone programming at the University of Nairobi back in 2007, 2008, and built this SMS blood bank system. A system that let rural nurses in Kenya text in what the day’s blood supply levels were and we built this visualization to show virtually in real-time what the blood supply levels were across Kenya, and more importantly, where blood was needed.

This all came about because I was working in the one building that had air conditioning. I was living out on the equator and in the one building that had air conditioning in this Village was this hospital. I was helping out with basically being the IT guy at the hospital in exchange for having an office with air conditioning. Basically, the second week I was there, I had these nurses come in to the office and say, “There’s been a traffic accident and we’ve run out of blood on our local blood bank we’re looking for volunteers to do emergency blood transfusions.”

I had a phobia with needles but when you’re in rural Africa and you to feel like there’s an emergency, you suck it up, and you donate the blood. You show off your bruise to anyone who will listen to your heroic story of how you really stepped up to a challenge. I did just that, and then about a month later, a different nurse came into the office saying, “Hey, there’s been an accident, we’ve run out of blood at our blood bank. We need to do a transfusion right now.”

Get more development news like this

By the fourth time this happened to me over the course of three or four months, out of my own self interest more than anything else, I started looking, what was going wrong with blood supply levels in this region the world? The short answer is it comes down to latency. There’s a guy in his pick-up truck that would be going back and forth, up and down the Canyon Coast and he would write down what the blood supply levels were in the hospital. Then the next time around if the blood supply was below a certain threshold, he’d come back in and replenish the blood.

We create a latency of almost four weeks from when you could have a shortage to when you get replenished. We thought we’d build this really clever system, this SMS blood bank system. Technically, it was bullet proof, it was rock solid but ultimately, it was deemed a failure. It didn’t work and it didn’t work because of just a fundamental misunderstanding on my part about the price of data in these markets.

The price of an SMS at that point in time represented a fairly substantial fraction of a rural nurse’s day’s wage. By asking them to send us data, we are essentially asking them to take a pay cut, something that just fundamentally wasn’t fair. At that point in time, I was in this really lucky position because of that one time, I had read and write access to the back-end billing systems of virtually every mobile carrier in East Africa because of the work I was doing at MIT.

I was in this unique position where I could basically manipulate the amount of money in a prepaid sub’s account. The whole genesis of this business was derived from essentially 20 lines of python where for every properly formatted text message with the day’s blood supply levels, we’d credit that rural nurse with about 10 Kenyan shillings, enough to cover the cost of the SMS and a penny to say thank you.

For that opportunity to earn one cent, we had virtually every one of these rural nurses re-engaging with our platform. It’s now a system that’s been deployed across East Africa. More importantly, a switch went off in the carriers head. Suddenly, Safaricom, the incumbent carrier in Kenya at the time saw this as a mechanism to increase ARPU, instead of squeezing more money out of that rural subscriber, they could start squeezing the Ministry of Health who was ultimately bankrolling this project. If they can squeeze the Ministry of Health, why not the World Health Organization, the Gates Foundation, P&G, Unilever, or Coca-Cola?

This represented a fundamentally new revenue stream for the operators and very quickly, the carriers that I was working with while at MIT were all hammering to have this this little platform built into their back-end billing systems too.

The short story is fast forwarding to today, the code base is far greater than 20 lines of python but it’s now been integrated to the back-end billing systems of 311 mobile carriers. We have operations now in 92 countries and we have the ability to essentially put money into the accounts of 4.56 billion prepaid subscribers, more than half of humanity. We can put money in their pockets instantly friction-free in denominations as low as 10 cents.

There is no company on the planet that has anything close to this capability. We are quite unique, which I think has been the reason why things like the fundraising has gone well for us. The challenge here is that now that we built the largest compensation platform ever, where’s the money coming from, right?

Paying For Free Internet(9:03)

I can talk a little bit about how we’re how are trying to address that question. The notion about about free internet is something that gets a lot of attention especially recently, whether it’s Project Loon where Google is floating balloons across Indonesia or Facebook besides buying drinks in about an hour or so next door, they’re also buying drones that are shooting lasers across Southern Africa. These initiatives are awesome. The challenge is that trying to figure out a system that really will scale to be able to truly provide a billion people with free internet. The cost of that is intense.

I think, it’s unlikely that a publicly traded company is really going to get their shareholders on board with paying out tens of billions of dollars a year to provide that level of connectivity which ultimately is the cost right now. If it can be done, the impact really is transformative. I mean living in Kenya for two years, I saw it myself. We’d see fisherman who would come in and they will be checking which marketplace had the best price for the fish, and then take their take their goods there.

You can see how conductivity really was and is increasingly transforming the lives fundamentally of people who frankly have a lifestyle that is nothing like any of you here in this room. If you think about the impact that your cellphone has on your day-to-day life. I would encourage you to think a little bit more broadly about how ultimately in a lot of these emerging markets as you give someone internet, you increase their income by anywhere between 15% to 30%. That really is profound especially when you’re starting to talk about people living below $2 a day.

The Price of Data(11:09)

I think, it is a worthy goal, it’s a very expensive goal but as these inexpensive Android handsets started coming to market, we are starting to see the ability now where people at least have the capability to get online. You can buy a cheap Android phone that’s almost now parody with the standard dumb phone that we associate with 15 years ago, the phone that plays Snake and sends text messages. You can buy a $20 Android handset now that does something very similar and that’s why you’re starting to see this massive adoption now of smartphones, not necessarily because smart phones are better but simply because they cost about the same amount and this growth is just going to continue right now in emerging markets.

There’s a little over 1.5 billion, people living in these emerging markets that are using smartphones. That’s going to be doubling over the next few years and if you start looking at data consumption, it’s going to go up by a factor of about six and this is problematic.

Hopefully, there’s a realization here that if data consumption goes up six- fold, the networks in these markets can’t support something like that. That becomes challenging and what the operators are doing in response to that are jacking up the price per megabyte. If you are making minimum wage in India, you have to work for 14 hours to get a one gig mobile data plan. In Indonesia, it’s 17 hours. If you are making minimum wage in Nigeria, you have to work for 27 hours to get a one gig of mobile data. In Brazil, it’s 33 hours. In Germany, it’s 30 minutes.

There’s this massive discrepancy here where people in these markets that could benefit the most from internet, it’s that the price discrepancy is such that it’s really difficult for them to at least consume as much internet as they like.

The Low Cost of Android(13:28)

The other thing that I think is worthy of noting is when we talk about smart phones in emerging markets, there really is only one offering system. We have tens of millions of users in a market like India and there aren’t over a few million iOS users in the whole country of 1.2 billion people. Who knows what the future holds for iOS market share an emerging markets. Right now, there’s only one game in town, and it’s Android.

I was in China recently and I got to see this handset factory that was printing out this brand name mobile phone during the day and then at night, there’s a change of staff, and then suddenly there’s another brand of phone that was being manufactured that was white labeled and costs 70% off of what the original phone was because it didn’t have the name brand. These phones now are just becoming increasingly fond of this.

People are able to start buying these low-cost Android phones but now, the challenge is they’re struggling to figure out how to actually connect them to the internet. If you look at the smartphone penetration in Brazil in terms of the Android share, the sad thing here is that more than half of all smartphones in Brazil are not connected to a cellular data plan. They’re being used like that Nokia phone that used to play Snake and send text messages. They’re not benefiting from the internet and this is not because people in Brazil don’t understand the benefits of the internet, this is because if you’re making minimum wage, you have to work for nearly 33 hours in order to get that mobile data plan. In my mind, that’s a travesty.

You’ve got tens of millions of people, in fact, hundreds of millions that are carrying around essentially this transformative super computer in their pocket and they’re not able to get access to the internet and that, in my mind, is something that, we collectively should be trying to figure out a solution to solve. That’s the reason why we built this technology.


Going back to this compensation platform, this platform that enables us to essentially credit a prepaid mobile sub’s account, the notion there is like we want to give people as much free connectivity as we can afford to. Who’s gonna fund the bill?

I think, as unpalatable as it may be for people on this audience, it is important to understand who actually is paying for drinks next door. It’s advertising. Advertising is paying for your drinks after the session. Advertising is the thing that ultimately is going to be funding the drones that are shooting lasers or the different initiatives. It is advertising that enables Google to organize the world’s information, and in my mind, it’s advertising that ultimately will be able to make the internet free.

There are just 300 billion dollars that are being spent on advertising in the developing world. That $300 is going into the pockets of the guys who own the television channels in Delhi. The billboards in Nigeria, the radio stations in San Paulo. This 300 billion dollars is funneling into the pockets of the 1% of the 1% of the 1%, the hyper hyper wealthy in these emerging markets.

It’s also funneling into traditional 1950’s style advertising channels and it’s because these advertisers don’t have a better option. I think that there is an opportunity here to be able to start figuring out how to redirect some of that spend away from the people who own those billboards or those TV stations and directly into the pockets of the very consumers that advertisers are trying to reach.

If we can do that, if we can even redirect 10% of that total spend of advertising in the developing world into people cellphones, we can provide one billion people with free internet on a sustainable basis. This is not something that we need, you need to talk about things like tens of billions of dollars in order to solve this problem but the money is there. I think, this problem can be solved. I don’t believe it’s going to be solved by large internet companies putting a philanthropic projects together. It’s going to be solved by commercially sustainable business models as well as CorTech, and that’s the thing that we’re excited about doing.

The New Users(18:50)

It’s not easy when you start thinking about developing for people in these markets, trying to think about what does Snapchat look like when you’re a Sri Lankan bus driver or a farmer in Ethiopia. It’s different. There are a lot of different factors.

The handsets that support these markets not only are most of them white or gray label OEM type handsets that are coming out of China, but a large fraction of them are quad-SIM which, is a novel thing for a lot of people in this market. Quad-SIM, meaning, you can put four SIM cards into it.

If you start talking to the average Android user that the Android user living in these emerging markets, it’s amazing how savvy they are about which SIM do I use at what period of time in terms of trading off like Tera Freight’s or calling someone on my Bharti Airtel SIM who also was on Bharti Airtel so we can to talk a lot cheaper. These users are extremely savvy in particular domains and they’re really a fun user base to build for.


I’m going to leave it at that. I’m happy to take questions but I also want to encourage you guys to start thinking about the impact that you potentially can have. Android is the fastest growing technology in human history. No technology has been adopted faster than the Android operating system, and in my mind, that’s less to do with how awesome Android is, but had more to do with the fact that Google made it such that these guys in these manufacturing plants in China are able to slap something onto the handsets that they’re pushing out.

The fact of the matter is that it really has the potential to have a profound impact on people’s lives. I think, that we have a shot here given the technology that we built, the momentum that we’ve cultivated, the team that we have. I think we have a shot it potentially pulling this off. We have a good shot at redirecting the amount of capital that’s currently going into the pockets of the people who own those billboards and redirecting it into the pockets of Android users in his emerging markets, which will enable us to provide a billion people free internet.

Ultimately, if you believe the world bank statistics, give them on average, an income lift of anywhere between 10% to 30%. You don’t get many opportunities in life to have an impact at this kind of scale, to have the potential to really have an impact on the lives of a billion people. We’re excited about making ours count.

If you are tired of working on that little email widget or whatever it is that you were working on that I’m sure is taking up a lot of time but may not necessarily have a profound impact on the lives of the billion people, please come talk to us. Thank you.


Do you think it’s worth it for advertisers to advertise to these users who can barely afford to pay for data?

I get that question a lot, you’re right. This is this is a good practice pitch. That’s the first question, they’re too poor to pay for data, what are they going to buy? Unilever is the second largest advertiser in the world. The majority of their revenue is coming from the developing world. When you start looking, and this is true actually across the board. If you ask, your it’s Google or Facebook, Pfizer and Merck, P&G, Unilever, Coca-Cola, or Pepsi, all of these companies know that their future earnings and revenue growth is not coming from this country, and certainly not coming from Western Europe. They’re all focused on emerging markets which is why emerging, like that 300 billion dollars that’s currently being spent in the developing world, is growing faster than any other market. They’re not buying iPhones, but they are there buying a hell lot of shampoo and a wide range of other standard consumer products.

You mentioned tapping into this 300 billion dollars, how exactly are you tapping into that?

We’re building an audience first and foremost. We’re building an audience of profiled, engaged, active users and then we’re making the case to advertisers that, “Hey, you should, instead of spending that money on that billboard, instead of going out and buying that TV ad, you should advertise through our app, we have an app that gives people internet.” It’s doing pretty traditional stuff like interstitials and banners. It’s getting people to try out other mobile apps and in practice. What this means in terms of engaging, we’re a company of 70 people but we’re almost all either data scientists or software developers. We don’t speak the advertising language but we just got a guy who runs publicist. He controls more of that 300 billion dollars spent than any other human on Earth. He runs a holding company that ultimately means that he introduced me to the CMO of Coca-Cola and literally, within 30 minutes, I had a flight we charted from Logan to Atlanta to come have me meet the Coke executive team. We have a technology, that can have a really transformative impact on brands if they become aware of it. We as a company need to do a better job of communicating that value proposition but when the brand’s get it, they get it instantly and, I think what we’ve seen is they’re able to start shifting spend and doing that in a big way.

How is this different from Free Basics?

Our whole company fiercely believes in the importance of net neutrality. At the same time, people get religious about this debate and I can see reasons to get religious upon this but the fact of the matter is, even before they were regulated away, we had far more users than Facebook did in India. People want free Facebook but they also want to do a Google search, they want to download an app, they want to watch a movie on YouTube, and so, they would use our platform as well. When we provide free internet, we’re not the gatekeeper of what content can be consumed and what content cannot be consumed. We’re essentially putting money into their account. That money can enable them to download whatever app they want, they can go to any website they want, they can do whatever they like on their phone. What that has meant is that we’re not concerned about the regulations associated that are happening right now in India, but more importantly, consumers are loving it. If you look at our penetration versus Free Basics in terms of the number of operators, the number of users, our model’s better.

In terms of scalability, what’s the next step?

We are laser focused on one thing and that one thing actually is not the 4.56 billion number. The one thing is related to that 30 million number that I alluded to the beginning of a talk. We gave 30 million people last year free internet. We are wedded to the idea of getting that number to one billion and that’s it. That is the thing that we need to do, and I think we have a chance of pulling it off. It’s going to take us at least another four years but we’re on that trajectory now and it’s leveraging this core infrastructure, this global infrastructure that we’ve built where we have the potential to provide free internet to 4.56 billion people and trying to actualize that for a quarter of them. We know we’re heads down focused on explicitly that, and explicitly trying to get the funds to cover that cost from advertisers who are increasingly looking to mobile as the way to engage with their next billion consumers.

You said we give the users a trial version of the app, we reimburse them for their time?

You can think of it like that. For example, Amazon is a big client of ours. What’s interesting in India is that people who make purchases on their phone in India, their average session length is anywhere between 90 to 120 seconds. What that means is they know what they want to buy, they open up the Amazon app, they type in the thing they wanna buy, the push buy and then close the app. Amazon as well as most e-commerce companies now makes money by getting people to browse, people who bought this also bought that, they want to get people to spend more time in their app. They want to create that habit of when you’re at the bus stop waiting for your bus, open up Amazon and browse. What we’re doing is for every 10 megabytes that you browse on Amazon, we’re going to give you 20 megabytes. That 20 megabytes can be used for anything. That is 20 megabytes, you can you can go to Flipkart, or you could do a Google search or go on Facebook or whatever. That’s the nominal notion for and sent the App Store.

You have mentioned Jana as app, as a platform, as infrastructure, what’s the infrastructure, what’s the IT stack that you have?

It’s a portfolio of apps. We have an App Store, and we’ve just launched a browser. All of this is being essentially enabled by what we call AirCom internally and that’s the core technology that gets integrated to the back-end billing systems of the mobile operators. It’s harder than you think to put a particular amount of money into a prepaid sub’s account. Even in India, the Bharti Airtel servers encounter brownouts and suddenly, you can’t actually deliver it in a particular way through going directly to the carriers. This is why this problem is hard is that when you tell someone, “Hey, I’m gonna give you 10 Rupees of free internet,” and they don’t receive the 10 Rupees, you incur wrath.

What we’ve had to do is build a lot of algorithms and ultimately, a lot of redundancy. In a lot of our major markets, we have at least six different channels to be able to credit someone, basically put money onto someone’s handset. We have an algorithm that constantly is looking at a condition on the time of day, condition on the volume that were pushing, condition on peak loads, we can figure out which channel is most reliable, and then have a fail over, retry, logic to ultimately get deliverability.

If we’re just going directly to the operators, were talking about maybe 85% deliverability but we can get to well over 95% and that’s technology that has evolved over now almost a decade but it enables us to do what we do. Hopefully, that answers your question. If it doesn’t, our VP of engineering is also here, we can we can get into the nitty-gritty of what the text that looks like.

I was wondering if you are considering the contents of the advertising, the quality of it, and possibly allowing your end users to have some sort of say in the ads that they see?

That is a great question and we’re not doing as good a job as I think I’d like to be doing. It’s a challenging model to try to figure out. Facebook has a show, hide ads like these. We also get push back on like, “Hey, do you really want to be selling sugar water? Do you want to start thinking about the ethical implications of pushing these types of products that may not all be soap and toothpaste?” I think those concerns have some validity as well but I also feel a lot of those concerns come from this almost colonialistic mentality of, “Oh, people in these markets aren’t quite ready yet.” They shouldn’t necessarily be exposed to all the things that we get exposed to because they’re not, that somehow feels like they might be taking advantage.

I was thinking that you could have helped Pepsi in the recent snafu with Kylie Jenner in that ad that they put out and also, if you’ve got a sports fan, maybe they wanna see more sports ads etc.*

That’s ultimately what we’re doing at least implicitly and that’s what ad networks do also implicitly. They start looking at what are the things that, websites that are being going to, what are the things that people are clicking on and start building an individualized model of that user to try to give them an ad that they’re most likely going to click on. We do exactly the same thing. It’s not particularly novel, I think we could do better.

There so many companies out there that are basically just lurking, following the footsteps of people online. They’re just like tracking your cookies and then they’re making money in this arbitrage. If they can increase the probability that you accidentally click on some crappy banner ad by a fraction and percent, they make money. Those models are bad and I hope will go away soon. At the end of the day, you learned far more if you get in front of the user instead of trying to track him behind. Getting in front of them and asking them, “Hey, what it is that you want or interested in? What are the types of ads you wanna see?” That ultimately is the path that were gonna go down.

Have you faced any opposition yet from the political parties or groups that are in power?

Here’s our strategy and we’re very open about it. We try to go into a country as hard as we can and create dependence amongst people. I was part of a group within Safaricom that was iterating on something called M-Pesa which is mobile money in Kenya. It’s a similar strategy where we basically launch this thing that would have been blocked by the Central Bank of Africa and we just launched it quickly, and it grew quickly. People started becoming dependent on it, their lives started improving, and by the time the centralized bank can come in to regulate it, it was too late, the population loved it. That’s our approach with offering people free internet is that we go in hard. We give people as much free internet as we can. We abide by the net neutrality regulations, and ultimately, the governments, for the most part, have been extremely supportive. At the end of the day, we’re improving the lives of their citizens, we’re actually adding revenue to the local carriers. We’re taking money out of the pockets of the guys who own the television channels. The guys who own the television channels don’t like who they are anyway, they know they want to be in a different business. Those are the enemies that we do make. It’s guys who are the traditional media buyers, but the governments have been, for the most part, very supportive.

Are there any restrictions you guys are placing on ads or servicing for looking into, how good the connectivity is?

That’s something that is critical. When we start talking about rich media ads, regardless of the price per megabyte, when you’re on a 2G connection, a rich media ad just sucks. What’s striking is we’re not the only ones to have that insight. As were starting to integrate more rich media ads, we’re working on our own Wi-Fi caching and, I haven’t talked about Wi-Fi at all but I think it’s definitely worth mentioning.

If you look at our users, 90% of them in these emerging markets have access to Wi-Fi at least once a week, and what that enables is caching. You can start caching while that handset is on Wi-Fi, you can get all sorts of content onto that device. That wouldn’t be possible on a 2G connection regardless of the price per megabyte. That’s what we’re doing for things like video ads and we’re not alone in that. There’s a company called AdColony that basically serves video ads only through caching and only sucks them down when their phones on Wi-Fi. Those types of models, I think, a lot of Western companies don’t pay attention to, but that is going to be one of the critical things is to get really good at being savvy around when the handset is on Wi-Fi versus cellular.

Thank you!

Next Up: Build Fully-Featured Apps that don't Need a Network Connection

General link arrow white

About the content

This talk was delivered live in April 2017 at Droidcon Boston. The video was recorded, produced, and transcribed by Realm, and is published here with the permission of the conference organizers.

Nathan Eagle

Nathan Eagle is the CEO and co-founder of Jana, the largest provider of free Internet in emerging markets. Born out of his time as a Fulbright professor in Kenya, Nathan began Jana from the insight that smartphones can be used not only for global communication, but also data sponsorship.

Nathan graduated from Stanford University with a B.S. in mechanical engineering and master’s degrees in management science and engineering and electrical engineering, and he was named a Technology Pioneer by the World Economic Forum in 2014.

4 design patterns for a RESTless mobile integration »