Facebook can do better than Free Basics

(and I hope they will)

It’s easy to read this as a purely nationalistic decision of one country against a big corporation. It’s read in the US (particularly on SV) as ‘they don’t want us to do good’. But there’s a whole lot of reasons why it came to this point. Facebook made decisions along the way that clearly show that they wanted to solve their own problem, not the poor Indian’s problems. Facebook could have done better — and still can.

First the facts: Internet access in India is relatively expensive. This is not exclusive to India, as many other developing countries face a similar situation. The high cost is due to a combination of unavoidable factors, like the need to build infrastructure, coupled with other purely political ones that boil down to a lack of competition.

(I’m not aware of the full details, but for what I have read and talked with Indians on Twitter, there are many rules to set up an ISP there; this leads to concentration in the hands of fewer companies and make harder to set up competitive services in hard-to-reach, poorer areas of the country).

There are numerous ways out of this situation, some easier and some harder. Facebook seems to have chosen the one that was easier for them, while solving their own problem: how to bring more people to Facebook.

Facebook’s Free Basics was proposed a partnership between Facebook itself and the local carriers, the same ones that are in some ways to blame for the high cost of Internet access in India. It’s a very convenient setup for the carriers themselves and that’s a big part of the reaction against Facebook.

Free Basics doesn’t mean that mobile coverage in India would increase or improve in quality. It just means that current poor coverage could be used to provide a subset of the Internet. This setup presents two strong risks:

1. Instead of fostering investment, the current setup could be used as an excuse for carriers to delay investment in infrastructure that is necessary to bring true Internet service to the regions served by Free Basics. These regions are probably not profitable today, and would hardly be a priority anyway if it was not for the requirements for universal service.
2. Worst of all, Free Basics builds a very effective moat for Facebook. It’s an anticompetitive device. It sets the expected price for Internet service at a very low range. Many potential customers will prefer to live with the limitations of Free Basics for a long time before being willing to spend any money with true Internet access. This makes the lives of competitive ISPs even harder.

We said before that there are many ways to handle India’s current situation with regards to universal Internet access. Free Basics was just one option that dealt with the problem by partnering with existing incumbents. Not the best way to do it, not one that goes for the root causes; it’s just the more convenient from both Facebook’s and carriers’s standpoint.

Here in Brazil, we lived a similar situation over the past decade. At the turn of the millennia, Brazilian government privatized state owned telcos and opened the local market for competition. For some five years there was a lot of investment as the big carriers ran for a land grab focused on the bigger markets. Broadband access became a reality in the richer regions of the country. But about 10 years ago, the process had almost stopped to a halt. Carriers started to consolidate. Telcos had spent too much money to outrace each other; there was a glut of infrastructure in some regions, but little investment was done elsewhere, just the minimum to meet universal service regulations.

It was against this backdrop that competitive ISPs started to pop everywhere. Local entrepreneurs learned to build cheap infrastructure to serve Internet to underserved communities. Some of those projects bordered on craziness, building long distance links over hundreds of km to bring Ethernet in locations like the Amazon Forest or the poorest region of Brazil, the savannah-like expansions of the Northeast ‘Caatinga’. Access speeds were low — in some case, you were lucky to have 1 Mbps — but it was great for a number of reasons:

1. Even if slow, it was fully neutral Internet access. It means that customers could freely choose which services to use.
2. Technical knowledge was spread to a number of young people far from the country’s biggest cities. Many of those people are now well paid professionals. The mere existence of local ISPs gave these people jobs, a place to learn.
3. A local ecosystem was fostered, with small companies teaching skills like how to build your webpage, or how to code a simple PHP application. As a comparison, this is simply not possible given the technical limitations with Free Basics.

Another important move at this time was the launch of the PTT.br project. Fostered by the local Internet registrar (NIC.br), and financed with the money collected from domain registrations and IP allocations, the PTT was the first truly open Internet Exchange Point in Brazil. After a slow start, the main PTT location in SP skyrocketed. Big content providers started peering there with anyone willing to connect. Local ISPs point out that something from 30% to 70% of their traffic now comes from the PTT, which translates in much lower costs.

Traffic growth over the past 5+ years at the SP PTT

Recent research tells that independent ISPs in Brazil now serve nearly 10% of the market — which is a feat in and of itself, considering that they are battling giant companies like Oi, América Móvil and Telefônica Vivo. Best of all, from the standpoint of the Free Basics project: this access largely covers the poorest and most underserved regions of the country.

If that’s all true, why is it so nearly universally ignored?

Even in Brazil, data on independent ISPs is hard to come by. There are three main reasons for this:

1. Market analysts love big numbers. These are the ones that bring attention. And best of all, being public companies, they’re already supplying analysts with the numbers. Doing research on hundreds of small companies is many orders of magnitude harder.
2. From a technical standpoint, it’s hard to take a fledging ISP seriously. In some cases the installations are a little more than exposed wires and improvised poles. If you have a telco background, it looks like a nightmare, a disaster waiting to happen. But considering the reality of the place, it works, and it couldn’t have cost a cent more. (Note in comparison how fond people are of the first Google’s datacenter, and you’ll note how prejudiced this viewpoint is).
3. Most independent ISPs are fiercely defensive of their operations. Many operations faced deathly competitive threats from the start. As a result, those ISPs don’t like to talk at all. It’s a secretive market. Attempts to dig more information only raise more distrust.

Note that most independent ISPs are private companies that aren’t required to publish any information. That’s one reason to suspect that the actual number of customers served by those companies could be even bigger than current research tells.

Closing this already long post, what could Facebook have done to avoid such backlash?

Facebook did something wonderful a few years ago. As the company was growing fast, they decided to build their own datacenter. But best of all, Facebook published the specs. The Open Compute Project is a fantastic resource for entrepreneurs that are looking to improve their computing operations with world class designs.

My tip to Facebook — if I’m entitled to any — is to pick a page of Open Compute and do the same with universal Internet access.

Do studies on how to solve the problem using local people, local resources, and the cheapest hardware possible.

Learn from other places (like Brazil’s poorest regions) that came up with their own solutions.

Develop software stacks to allow independent ISPs to provide better services with less hassle. Provide VM images with DNS and IP address management tools. Provide user management tools. Provide logging tools.

Work with governments and regulation agencies about the importance of competition in the Internet access market. Help clear the way for small companies to be legally started.

Help to setup peering locations where independent ISPs can connect to each other and to the larger content providers.

Publish this information for free. Let the world learn. Let the local people understand that they can do something for very little, all the while learning more, acquiring new skills, and being truly inserted in the modern Internet world.