Rethinking Affordable Access

Everyone needs affordable access to communication. Affordable access strategies that don’t target everyone end up magnifying the digital divide in a kind of Matthew effect: those with affordable phone/internet services have access to ever increasing education resources, opportunities, services, social safety nets (the list goes on) such that the unconnected fall further behind just by standing still. You may not agree completely with this statement but let’s assume for the purpose of this article that it is true.

And if it is true, then we have a problem because access growth is slowing down. Growth is slowing whether you look at mobile subscribers or internet penetration. To date, a combination of public and private investments in telecommunications has managed to connect about half the world to the internet. Mobile Network Operators (MNOs) in particular are the runaway success story when it comes to providing affordable access in regions thought unreachable. So why, given the rapid growth of the last fifteen years, are things slowing down? The one-line answer is that we have connected the easy-to-connect half, the relatively wealthy living in relatively densely populated areas. Connecting poor people is sparsely populated rural areas is a much bigger challenge. In a recent Guardian article, Sonia Jorge of the Alliance for Affordable Internet Access (A4AI) suggests that universal internet access is as far away as 2050.

That same article suggests that the reason growth is slowing is a combination of lack of skills and lack of investment. It is undeniable that lack of skills to take advantage of digital tools and services is a significant barrier to uptake. It is also true that those skills are connected to affordability. We discover value online by experiencing it and by experimenting to discover what is valuable to us. When access is expensive, the discovery of value in being online is overshadowed by concerns about the cost of being online.

Lack of investment is also a complicated issue. The vast majority of MNOs are private companies that serve their shareholders. Even where MNOs are government-owned, there is often more interest in generating a profit than in providing universal, affordable access. It is unsurprising then that MNOs invest in areas where both population density and incomes levels are higher. Spectrum auctions that levy millions of dollars from MNOs in advance of network deployment only compound this trend as operators worry about an increased debt burden.

Universal Service Funds (USFs) are the primary mechanism that has been established to try to ensure that poorer and more remote regions also have access. Operators are typically taxed a percentage of their turnover that goes into a fund run by a Universal Service Agency, whose purpose is to finance access into unserved regions. The history of USFs and the associated agencies that manage them is pretty mixed. These agencies are typically government bureaucracies that have little experience managing millions of dollars and are risk averse by the very nature of being bureaucracies. Worse, funds with millions of dollars often attracts the worst kind of kleptocratic behaviour.

That said, USFs have made a difference. In the early days of mobile network growth, they created incentives for MNOs to invest in areas that were not economic priorities for them. This led to the growth of infrastructure into previously unserved regions. However, these subsidies have typically succeeded in regions where MNOs would have found themselves eventually, that is, areas that have populations dense enough to create profits for MNOs

Where USFs have really run into problems is in trying to push MNOs into regions that are not economically viable for them. Stories abound, although seldom publicly, of USF funds that have completely subsidised the capital costs of cell tower build-outs for MNOs in remote areas but those cell towers are not operated by the MNOs because the towers don’t generate profits, or perhaps not sufficient profit. In one Southern African country, the USF agency has paid for the build-out of 103 base stations in rural areas; 77 of which are not operational. This is not particularly surprising. Often, the operational costs (e.g. maintenance, supply of diesel for back-up generators, etc.) can exceed the revenues generated by calls on the base stations in remote, sparsely populated regions where income levels are lower than in urban areas. For an MNO that is accountable to its shareholders, it only makes sense to turn those towers off.

And that is where we find ourselves today. Access growth is slowing because the economics just don’t make sense. So what to do? A recently published report commissioned by the World Bank entitled “Innovative Business Models For Expanding Fiber-optic Networks And Closing The Access Gaps” attempts to answer that question. Their conclusions don’t exactly break any new ground for the World Bank, however. An extremely short (perhaps unfair) summary would be:

  • Enable market forces; the private sector has been most successful at deploying networks. Try to get out of their way.
  • Regulate only as a last resort where there is absolutely clear evidence that the market has failed.
  • Government investment should be cautious and only implemented with a clear vision.

The problem with the above recommendations is that this course of action got us to where we are now — but it doesn’t address the fundamental problem of access economics that the MNOs face. Nor does does the proposed strategy mitigate any of the downsides of MNO network deployments, which is often a lack of market competition. If new thinking is required, what might that look like? Here are some news ways of looking at the problem.

We Don’t Live in a Single Economy

The conventional analysis of the access challenge treats the economy as a single entity but the reality is that the economics of the rural poor is very different from comparatively wealthy big cities. French economic historian, Fernand Braudel, makes a distinction between global and local economies. Companies that operate in the global economy are easy to spot because the endgame for those companies is monopoly. Whether you are MTN or Amazon or McDonalds, the goal is to occupy the maximum possible amount of the market. These kinds of companies get most of our attention. Growing to this size is seen as a virtue. And indeed, there has been a lot of success in the growth of mobile access.

Silicon Valley has a particular obsession with scale. Where else would growth hacking be an actual profession. While there has been some pushback to this, scale remains the brass ring for internet companies in particular.

In contrast, local companies seek to serve an immediate market, one which they typically have much deeper knowledge of and are able to serve in a manner that a global company cannot. The economics of these local companies are different than those of global companies. Some things are more expensive, but many things are cheaper when it comes to serving a local market. Thus the local baker’s labour costs per loaf of bread are higher but they don’t incur any shipping costs and can tailor their production specifically to demand. Not to mention the fact that they can greet you every day and be a part of your life and your community.

This is not news. Governments around the world have recognised the importance of small and medium-sized enterprises (SMEs) and have policies developed to encourage their growth. The economic contribution of small business is substantial. In the United States, small businesses employ 47.5 percent of all private sector employees. In South Africa, SMEs contribute 34 percent towards Gross Domestic Product (GDP) and provide employment to about 60 percent of the labour force.

But nowhere do we see a deliberate strategy to empower small and medium telecommunication/internet access providers. Again, this is not surprising. Historically, building a telecommunications company required millions of dollars of investment in international connectivity, national network infrastructure, last mile infrastructure, handsets, agent networks, marketing, the list goes on. It was a risky bet too. Choose the wrong underlying technology (e.g., WiMAX mobile ten years ago) and you could lose millions.

The next big thing will be a lot of small things

All this has changed. Telecommunications networks are becoming disaggregated. Increasingly international, national, and metropolitan network infrastructures are available as wholesale services to any operators, lowering the bar to market entry for smaller operators who can focus on last-mile delivery.

Also, communication technology in general has dramatically dropped (and continues to drop) in price. Not only has WiFi technology become extremely affordable but all kinds of communication technologies from point-to-point microwave to GSM to LTE base stations, even fibre optics now have prices that are within the reach of the community network and small-scale operator.

“When a flower doesn’t bloom, you fix the environment in which it grows, not the flower.”

Alexander Den Heijer

Tell me if this sounds familiar: You’re in a meeting about affordable internet access and someone mentions the importance of community networks and small-scale operators. Someone stands up, usually from a big operator, sometimes government, sometimes the regulator, or even the ITU and says, “Oh we recognise the value of community networks… but do they really scale?” The implied answer is “No, of course not.” But what if we could connect millions of people affordably through small businesses i.e. ISPs and community-owned networks without scaling?

Not far from where I now live in rural Nova Scotia, is a small community of about 500 people called Lawrencetown. They got fed up with not having decent access to broadband and being ignored by the incumbent telcos and formed a cooperative to build a WiFi network that serves the community. They started out charging about USD45 per month for 20 Mbps service but reduced the price to USD30 per month because they found they were making too much money. They don’t need to be any bigger than they are. They were able to do this because:

  • Access to backhaul is affordable and open;
  • The administrative barriers to setting up an ISP are very low; and,
  • WiFi access technology is very cheap.

Their success has emboldened the community to explore the use of a cooperative to deliver both healthcare and energy.

We now live in world that has conspired to make it possible for local entrepreneurs and/or community groups to solve their own access challenges. What is missing is the enabling policies and regulation to let a million of these little (and not so little) flowers bloom.

Community-Owned Infrastructure

In addition to creating an enabling environment for local access solutions, we need to break out of the Manichean world of private sector networks vs government networks. In particular, community ownership of telecommunications infrastructure needs to be elevated to the mainstream of affordable access debates.

We need to dispel the picture that many people seem to have of community-owned networks: two tin cans and a piece of string. It is true that in the early days of wireless hacking with WiFi, tin cans actually featured but it was that kind of hacking that spurred the growth of WiFi as a credible backhaul technology.

More to the point, as the status of telecommunications infrastructure is changing from luxury to essential utility, there are lessons we can learn from the growth of another essential utility that started out as a luxury: the electricity grid in the United States. In the early part of the 20th century, much of rural America had no access to electrical power. Rural farmers and ranchers banded together to form cooperatives to build their own power infrastructure. But they likely would not have succeeded had there not been enabling policies and programs put in place by the government. The following 3 minute video which tells the story of US electrical cooperatives is well worth watching.

Today, those same cooperatives are rolling out fibre optic infrastructure to address lack of broadband service in those same regions. Cooperatives can take many forms. They can be tiny like the one in Lawrencetown. They can address access challenges in regions that aren’t viable for large operators such as Rhizomatica are doing in rural Oaxaca, Mexico, or Zenzeleni in the Eastern Cape of South Africa. They can embody commons principles established by Elinor Ostrom as are doing in Catalunya, Spain. They can enable farmers to dig their own fibre trenches to deliver 1Gbps (symmetric) internet service in rural northern England as B4RN are doing.

Cooperatives are a time-tested model for addressing collective resource challenges, especially but in no way limited to in rural areas. Why should we prioritise their growth? Here are a few reasons:

  • Cooperatives lower consumer costs. Cooperatives are non-profits and reinvest surplus back into the network, either by lowering costs (as Lawrencetown did) or by expanding the network. This is particularly relevant in areas that may have enough demand to support one network operators but not two. Subsidising a commercial operator to build a network may solve the access challenge but does not address the affordability challenge.
  • Cooperatives invest in local economies. In a remote community, calling someone a block away typically generates revenue for the incumbent operator headquartered in the capital. It siphons revenue away from remote, poor communities. Community-owned networks ensure that revenue spent on telecommunications is spent locally, creating more local economic circulation.
  • Cooperatives build local skills.  The technical skills to build and manage communication networks is within reach of anyone with the right learning environment. Cooperatives provide a space for local skills development in a 21st century career.
  • Cooperatives can help to unwind inequitable social systems. We know that communication technology is an amplifier which can magnify positive outcomes like entrepreneurship and education but which can also reinforce existing social and economic inequities. Because cooperatives are not driven by a profit motive and because they are local, they can both understand and collectively choose to address challenges that face their communities in a meaningful way.
  • Cooperatives build… community.
    • Perhaps that sounds like a tautology but there is nothing like having a project in common to build community bonds/ties and the social fabric that holds us all together.

This is a “yes, and” scenario

This article should not be interpreted as an argument against big telcos. The success of large-scale, commercial telecommunications networks is undeniable. They have great advantages thanks to their economies of scale, however, the evidence is mounting that the large network model alone is not enough to connect everyone affordably. Small businesses are now are practical reality in the world of telecommunications and the internet, but are limited, not by technology, but by the lack of an enabling policy and regulatory environment. Similarly, community ownership of telecommunications infrastructure in the form of cooperatives and similar structures is a proven success model that cries out for policy, regulatory, and investment support. Locally-owned commercial and/or community connectivity initiatives will grow the market which will ultimately be good for big telcos too.

The best news for regulators is that enabling local connectivity initiatives directly addresses one of the biggest concerns raised in the World Bank report: risk. Smaller initiatives can fail without causing systemic harm. They can innovate faster and be more responsive to local needs. Cooperatives are owned by their members. What better way to ensure skin-in-the-game, arguably one of the best ways to mitigate risk.

It’s time to move the word “local” into the mainstream of telecommunications policy, regulation, and investment.

The banner photo is of the amazing flowers of Namaqualand in the Northern Cape of South Africa that bloom annually in the desert when right conditions occur. I couldn’t think of a better image to illustrate what can happen when you create the right conditions. Image courtesy Martin Heigan.

Fibre Feudalism

Rapidly growing demand for broadband is changing the landscape of African telecommunications infrastructure.  This change has been enabled by the spread of both undersea and terrestrial fibre optic infrastructure, as well as the proliferation of low-cost smartphones, which can deliver broadband through either 3G/4G services or WiFi.  In this new world, two critical choke-points emerge as key barriers to the growth of broadband on the continent.  One is access to wireless spectrum through which last mile services can be delivered and the second is access to fibre optic backbones.  I’ll deal with the former in a future post but in this article I want to concentrate on fibre backbones and the critical role they play in enabling a future where everyone has affordable access to high-speed internet.

The impact of the arrival of high-capacity undersea fibre cables to the continent in 2009 cannot be overstated.  It is as significant a technological change as the introduction of mobile networks back in 1994.  The mobile revolution was easy to recognise.  Everywhere you went, you saw someone with a phone; the impact was right before your eyes.  Fibre optic infrastructure, on the other hand, is harder to recognise because most of it is invisible to the casual observer.  Yet the impact is profound.  There is a direct causal link between the arrival of undersea cables to the development of what is now hundreds of thousands of kilometres of terrestrial fibre optic backbones across the continent, and from that to the rapid spread of Fibre To The Home (FTTH) services in African cities.

This infrastructure enables existing operators to deliver mobile broadband services more effectively and scalably to their customers but, perhaps more importantly, it also enables a whole new generation of service providers from FTTH operators to WiFi network operators to new market entrants with licensed spectrum.  The relatively infinite capacity of these fibre optic cables means that they are what economists refer to as a non-rival asset, that is to say that use of the capacity by one operator in no way affects or impedes the use of it by another.  On any modern fibre infrastructure there is enough room for all players using all the capacity they can for the foreseeable future without encroaching on each other.  And should more capacity be necessary, fibre is also relatively inexpensive to upgrade compared to other kinds of telecommunication infrastructure, often only requiring the equipment lighting the fibre to be upgraded at the termination points.

Sadly this picture of digital abundance does not reflect the current reality.  What exists now is an artificial scarcity created by fibre optic network owners that bears some resemblance to medieval feudalism.  Those who own fibre network infrastructure use it to extract high rents from those who don’t own any.  If you are a broadband service provider and you don’t own fibre infrastructure, you are completely at the mercy of your upstream fibre backbone provider who, because there are few backbones where there are multiple fibre routes, is at liberty to charge whatever the market will bear for access.  As the cost of last mile technology (whether wireless or FTTH) steadily decreases, the cost of backhaul becomes the critical cost factor for small and even large service providers.

Yet by owning even a modest amount of fibre infrastructure, the game changes.  Fibre network operators routinely engage in capacity swaps with other fibre operators.  Operator A owns an FTTH network in a small city.  Operator B owns a backhaul network in the same country.  Operator B would like access to Operator A’s customers and Operator A would like backhaul capacity to the coast or beyond.  A deal is done. Or Operator C owns capacity on one international submarine cable and Operator D on another.  A capacity swap is arranged so that both operators now have redundant international fibre connections. Each operator can share the capacity without it affecting their own capacity requirements.

These capacity swaps are happening everywhere where fibre exists on the continent.  I liken this to medieval aristocracy marrying off their sons and daughters into other aristocratic families in order to strengthen their hold on their titles.  For those who don’t own fibre, they are the serfs and vassals who may toil as hard as they may but ultimately serve at the pleasure of whoever provides their upstream capacity.  The metaphor isn’t by any means perfect but it is a powerful illustration of  what I believe to be a fundamental roadblock to broadband development.

I think we can all agree that feudalism was not the system of governance that best serves the majority.  But how did we get here?  In broad strokes there are privately-owned networks and state-owned networks.  The majority of privately-owned networks belong to the large network operators like MTN, Orange, Vodafone, Airtel, and others.  They engage in this feudalistic approach because this has been their standard approach for many years.  They seek to exclude others from the market.  For many years this has been through actively inhibiting the release of wireless spectrum to anyone but themselves but more recently, where new operators have gotten around the wireless spectrum barrier by using unlicensed spectrum or by building FTTH networks, it has been through extracting high rents for access to their fibre networks.  A recent report by Xalam Analytics, highlights this very issue.

“…fibre infrastructure is not only a source of competitive advantage for Tier-1 MNOs; it’s a potent weapon to drive smaller providers out of business, by keeping their fibre input costs high, while slashing average revenue.”  ~Xalam Analytics

For state-owned networks, prices remain high for two reasons.  One, the government has usually borrowed a large amount of money to build the network. Sadly this may often be several times what they should have paid for the network due to vendor influence and a lack of technical expertise on their side to back them up during negotiations.  Whichever the case, governments are usually keen to get rid of the debt burden and thus not averse to matching the private sector in the high prices they charge.  The second reason is that any attempt to reduce prices below what the private sector is charging is likely to be met with legal challenges from private operators charging that the government is undermining their sustainability.

This thinking by both governments and the private sector is bad news for broadband development on the continent.  There is a need to stop thinking about fibre optic networks as commodities in which the market price is determined by controlling supply and demand and to start thinking about fibre in the same way that we think about roads, railways, shipping ports, that is to say as infrastructure.  The strategic goal of infrastructure is not to derive economic benefit from the asset itself but to generate economic benefit by maximising the use of the asset.  The more drivers we have on roads, the more economic activity is generated.  It is true that toll roads exist but the reason for their existence is typically related to managing congestion and, as pointed out earlier, fibre optic networks are unlikely to suffer from this problem.

How do we get to a fibre infrastructure driven utopia?  How do we escape from the feudalistic system we have found ourselves in?  I have previously proposed a strategy for state-owned networks that would see part of the network sold off to a consortium of smaller operators.  I remain convinced that, for many countries, this is exactly the strategy that should be pursued.

For privately-owned networks, I have a different proposal.  In Mexico, as part of their license, satellite operators are required to donate a percentage of their capacity to the Mexican government, on the condition that it is used for security, public safety, universal service, public sites or other official uses.  Universal service is the key phrase there.  What if governments gave operators the choice of contributing to a universal service fund or making some small percentage of their fibre capacity available to universal service activities.  This could be a game changer for community networks and small-scale operators wishing to provide access in underserved areas.  This would have limitations as there is a catch-22 that underserved areas are unlikely to be served by fibre but microwave networks can extend fibre access by hundreds of kilometres.

Of course this does beg the question of financing.  If the cost of access to fibre is lowered, operators may argue that they will not earn a sufficient return to justify investment in fibre networks.  There are a few sides to this issue.  One is the economics of scarcity vs the economics of abundance.   High fibre costs restrict the number of people willing to pay for access.  Low fibre costs are likely to drive up volume where existing purchasers simply increase their volume without increasing their costs and new market entrants become new purchasers of capacity because the lower costs enables their business models.  There is a kind of prisoner’s dilemma with fibre at the moment where operators choose to undermine each other rather than acting in the collective interest so that everyone wins.  In some cases, layer separation, where fibre networks are restricted to the provision of wholesale-only services may improve the situation.  There is some evidence that wholesale fibre initiatives like CSquared (formerly Project Link) are having this kind of impact in the greater Kampala region in Uganda but the situation is less optimistic looking at the wholesale network built by Korea Telecom in Rwanda on behalf of the government.  It seems clear that there is little point in implementing a wholesale network strategy without having a clear strategy for addressing the pricing of the network.

One strategy might be to have a state-owned wholesale network always match the very lowest price from the private sector in the most competitive part of the country.  For example, if the best price for fibre capacity from Nairobi to Mombasa is $5/Mbps, then the state-owned, National Optic Fibre Backbone (NOFBI) network must offer capacity anywhere on its network at $5/Mbps.  A similar situation in Uganda where the best price on capacity from Kampala to Entebbe must be matched everywhere on the government’s National Backbone Infrastructure (NBI) network.  This would have a catalytic effect on service delivery to underserved regions.

Perhaps the most interesting and radical notion would to treat fibre backhaul networks as a “common pool resource” based on the principals of the commons articulated by Nobel laureate Elinor Ostrom.  Under such a system, fibre network owners that join a common pool resource agreement, consent to give up ownership of the asset to the commons but are compensated in an ongoing manner in proportion to its use and maintenance costs.  Each network that joins the commons, gains access to the entire network and increases its size and scope.  Ensuring that all members of the commons contribute and/or are compensated fairly for their contribution requires careful rules and accounting but the result is something that delivers maximum fibre capacity to all at the lowest possible cost. in the Catalunya region of Spain operates a growing FTTH fibre optic network as a wholesale, common pool resource that now serves over 5000 households.

One way or another, there is a critical need to unlock the vast capacity lying dormant in most fibre infrastructure across the continent if African countries are to realise their full digital potential in connecting everyone to affordable broadband.

I am grateful to my colleagues from Rhizomatica for invaluable feedback that improved this.  Any errors are all mine.


The Internet Is U-Shaped

When we think about the problem of achieving affordable access to the Internet for all, the discussion often focuses on broadband targets.  These targets are moving goalposts as infrastructure improves.  Broadband used to be defined at 256Kbps, now it might be 2Mbps or 5Mbps or something else depending on who you talk to.  Nobody wants to be seen as providing second-class broadband to their country.  And yet, the higher we set our targets for broadband, the more ambitious those goals become.  This is particularly true when we are talking about rural regions with variable population distributions with low ability to pay.  I have come to wonder whether simple broadband targets are the right strategy.  Perhaps we need something more nuanced.

First I want to suggest that slow Internet is still extremely valuable Internet.  Only very modest speeds are required for the basic operation of enormously population messaging services like WhatsApp, WeChat, FB Messenger, et al.  The growth of those services in the last few years has been nothing short of phenomenal.  It is with that in mind that I along with Steve Esselaar and Christoph Stork have been pitching the value of making 2G Internet access available for free to anyone with a data-capable phone.  You can read more about that idea at Zero-Rating: A Modest Proposal or watch the 10 minute pitch I made to the Mozilla Equal Rating Challenge last week.

It is an interesting thought experiment to ask yourself which is the more equitable strategy: connecting 80% of the population at 5Mbps or 100% of the population at 2G speeds.  The answer of course is that it is not an either/or question but that both strategies might be important.

And of course, 2G, while essential and extremely valuable, is not enough.  That got me thinking about where the value really exists on the Internet.  And for me the other place of high value exists at either end of the broadband spectrum.  I value small amounts of bandwidth for texting, emailing, etc and I value video, more specifically I really like YouTube.  I don’t like singling out a particular platform but no one else is delivering value the way YouTube is at the moment.  And at the low-end of the value proposition somewhere are top-heavy news sites like CNN.  I don’t mean to pick on CNN, there are hundreds like them; sites that deliver 20KB of reading material while taking up 5MB of capacity.

A word about YouTube.  I think YouTube is one of the pedagogical marvels of the decade.  There is virtually nothing you can’t learn to do on YouTube whether it is learning to be a carpenter, to fix your car, to write music, to be a magician, to fix electronics, to dance, to understand particle physics, the list is endless.  Chris Andersen of TED understood this six years ago in his talk on how YouTube is driving innovation. But it has gotten even better since then. Nadiya Hussein, who won The Great British Bake-off in 2015 is credited with learning to cook via YouTube videos.  Examples like this are legion. In communities where no affordable broadband exists, people travel to areas with free or cheap WiFi just to download YouTube how-to videos to take back to their communities.  Some even take requests.  I see this as well in my sons who are between ages 11 and 12.  They acquire skills on YouTube that frankly amaze me and they put more effort into learning there than they do in school because they are tapping into what they are passionate to learn about.  YouTube is also valuable because it bypasses a lot of local language and literacy challenges by allowing people to express themselves directly.  It’s not a panacea in this regard but it is a real vehicle for local expression.  Again I want to emphasise that it is not YouTube in particular that I am advocating for but the value that YouTube is providing right now.  What it might be in the future is another thing entirely.  Platform diversity is just as important as content diversity.

So if I can’t have the whole Internet, or if I have to decide what I get first, I would choose slow and fast Internet. My graph of Internet value vs speed looks something like this one on the right.  If my value framework is true, does that tell us something different about how to prioritise the development of Internet infrastructure?  For my it places a clear emphasis on getting everyone at least 2G access to the Internet but what does it mean we should do at the other end?  Conventional wisdom says invest in more broadband networks, 3G, 4G, 5G and so on.  In the last year WiFi networks have come to challenge that orthodoxy in Sub-Saharan Africa but, with a few notable exceptions, they tend to thrive in urban areas where there is existing high-speed backhaul infrastructure.  Getting YouTube or Vimeo or other video platforms to sparsely populated rural areas in an affordable manner is a real challenge.

That is why I am so excited about a pilot project that Google have in the Philippines called Google Accelerator.  Yes, not the catchiest name in the world but still an amazing project.  Google Accelerator is what I would call a micro Content Distribution Network (CDN).  You’ve heard of Akamai, CloudFlare, Google Global Cache and other CDNs that work at the network operator level and provide national-level caching for petabytes of content.  A micro-CDN works at the small organisation level.  Imagine a TV set-top box that is WiFi-enabled.  Now picture it with a 4 terabyte drive that is packed full of YouTube videos.  Anyone who connects to the WiFi access point of the set-top box gets instant high-speed access to YouTube.  Suddenly the high-value end of the Internet is available anywhere.  Of course the videos need to be regularly updated and that is challenge.  The ingenious part of Google Accelerator is that the content is refreshed via DVB satellite.  Pretty cool.  Google Accelerate is also remarkable because the solution is very affordable.  The components are commodity devices that might cost a few hundred dollars in total.  The challenge lies more in the refresh of the devices and what that costs.

Google aren’t the only ones who get the value of micro-CDNs.  Kenyan startup BRCK also seem to have pivoted to a content-based strategy in which their latest WiFi device the SupaBRCK which also has a hard drive specifically for local caching of Internet content.  This doesn’t mean there aren’t challenges.  Platforms like YouTube and others can be particular about allowing anyone else to cache their content.  Coming up with a licensing strategy that allows for independent micro-CDNs may be necessary.  Also, the trend to HTTPS encryption for all websites is a factor.  HTTPS encryption is great for blocking casual surveillance by bad actors but presents challenges for CDNs that can’t see what content is being browsed in order to cache it.  Large CDNs like Cloudflare solve this by decrypting and re-encrypting traffic as it passes through them but it is likely not the simple for micro-CDNs.

So what do you think?  Is the Internet U-shaped in terms of value vs speed?  Does that suggest different ways of prioritising infrastructure development?  This article was written partly as a provocation to test this idea that has been growing on me for a while. I’d love to hear rebuttals, affirmations, and/or alternative theories.



Failure to TED

ted_rediscover_this_dayRecently a bot informed me that it was the 2 year anniversary of my speaking at TED Global 2014.  And that reminded me of what I have been putting off writing for a while, putting off well, because I find it hard.

Shortly after TED Global, I wrote about my experience of the event. What I didn’t mention was the talk that Chris Anderson gives speakers prior to going on stage about the post-talk process for putting TED talks up online.  He explained that not all talks go up immediately, that it might take up to a year for that process to come to fruition.  He went on to say that, for some talks, they might decide that the speaker had better talks to give.  I can’t remember exactly how he phrased it; it was delicately put. In a nutshell, not all talks go online.

Obviously this is a big deal. It was a primary motivating factor in my decision to agree to do the talk. By this point, you have guessed where this is going, my talk wasn’t selected for publication.  This was a pretty big disappointment.  I really hoped that having the talk online would be a powerful tool in spreading a different narrative about affordable access.  What was emotionally worse though was having to respond to friends and colleagues when they asked where they could watch the talk online.

For the first year, I was able to say that it might be in the queue.  Due to a communications mix-up they failed to inform me of their decision and, pusillanimously, I preferred to assume that is was still in the queue rather than ask the TED folk the question outright.  As the months wore on I began to suspect the answer.  After 11 months, I steeled myself to ask and Bruno explained in the nicest way possible that while the talk had worked in the theatre, it didn’t “land” as a video.  Once again they had a delicate turn of phrase in letting you down.  I guess they’ve had practice.  It was still a very hard email to read.  Hard to square the reality that was with the reality I wanted.

But I like to imagine that I am not the person I was, someone who would have internalized this failure and hugged it quietly to myself while accepting as a some kind of judgement.  Writing this is part of proving that to myself and I can point to two things that helped me.  One is Carol Dweck’s Mindset, a book I wish I had read when I was an adolescent.  Mindset is one of the most important books I have read as an adult.  It helped me to recognise that failure is not a personal judgement on my worth as a human being but an opportunity to evolve.  It started a journey that I am still in the midst of.  Sounds obvious doesn’t it?  Easy to say, hard to live.  I am still internalizing the lessons of the book and losing my hang-up about failure is a work in progress.

The other inspiration for writing this is Princeton professor, Johannes Haushofer, who earlier this year published his CV of failures in which he catalogues in Curriculum Vitae style, the jobs, fellowships, programs that he has applied for but did not get.

Haushofer himself acknowledges he was inspired an article in Nature by Melanie Stefan.  I want to say thanks in public to both of them for giving me courage.  Their failure CVs emphasised for me the importance of owning failures (without dwelling on them) as a means of internalising learning and moving on.  What I also learned was the value of doing it publicly.  Social media has become a place where we are encouraged to demonstrate how awesome we are; the cool places we’ve been, the fabulous things we’ve eaten, the famous people we’ve met.  It fosters a sense that everyone else is doing better than you are, having a better time, being more successful, etc.  We don’t often reveal when we fall and it hurts.  Haushofer’s CV was an inspiration in that regard although interestingly although in his CV the emotion is all implied.

So what did I learn from my failure to TED?  A few things.

  • The right offer at the wrong time may not be the right offer.  The offer to speak at TED, which came via a very generous recommendation, came about two years too late or maybe a little too early.  The organisers wanted me to talk about Village Telco but by that time I already knew that Village Telco was not going to take over the world.  It is still going well as an Open Source project (thank you for asking) but it is not the phenomenon I once hoped it might be.  Thus I wanted to give a different talk than they were expecting, which was fine except that it led to mixed communication in the preparation of the talk and mixed result.
  • It’s all about the story. I had important stuff to say but I struggled to weave it into a cohesive narrative.  We are storytelling creatures.  I know this but somehow lost the plot in getting there.
  • When preparing start actually speaking your talk as soon as possible.  I spent too much time writing things out and re-writing.  It would have been better to spend more time on delivery so that I was less nervous about forgetting on the day.
  • Don’t hesitate to clear up uncertainty immediately.  Letting things hang on is always worse in the end.
  • Most importantly I learned (am learning) to fall down in public and get up again.

It was still a cool experience and am grateful for the opportunity.  Would I do it again?  Of course.   I enjoy public speaking and sometimes I do it moderately well. Wish it hadn’t taken two years to write this but getting it out there at all is a step forward.


Africa Telecoms Infrastructure in 2015

This entry is part 2 of 5 in the series Annual Review

Welcome to the 2nd annual review of telecommunications infrastructure development in Africa.  Last year was my first attempt at painting a big picture of communication infrastructure development on the continent and it seemed to fill a gap in public information.  Also I found the process very useful personally.  I find I don’t stop to take stock often enough.  This year’s review is a little more ambitious.  Below you will find links to over 140 news articles covering everything from fibre to spectrum and more.

Undersea Cables

If 2014 was the year that a number of “powerpoint” undersea cables projects disappeared into the digital ether, 2015 is the year when I learned to stop making predictions about undersea fibre capacity.  Only a month ago I was predicting the end of new African undersea fibre projects — only to be wrong-footed two weeks ago by Liquid Telecom’s announcement of a brand new cable called LiquidSea to be built up the east coast of Africa. Indeed things have  been busy in 2015.

The ACE cable consortium announced that they would be extending their cable from Sao Tome all the way to South Africa.  They also added new partner countries, including  Benin, the Canary Islands, and Cameroon.   Cameroon in particular was notable this year because not only did they join the ACE consortium but they also announced an agreement with MainOne to build an extension, called the Nigerian Cameroun Submarine Cable System (NCSCS), to connect Cameroon to their undersea cable.  As if that weren’t enough, the Cameroon-Brazil Cable System (CBCS) was announced in October which proposes to connect Cameroon with Brazil.  Given that Cameroon only announced its first 3G services about a year ago, this is a remarkable change of events.  Time will tell whether it leads to a more competitive market there.

That wasn’t the only cable connecting Africa to Brazil.  The EulaLink cable is planned to connect Brazil with Portugal and has potential landing points in Cape Verde and the Canary Islands.

Existing undersea cable operators were not standing still either:

Other smaller announcements included Algeria’s plans to build an undersea cable connecting to Spain.  Angola also announced an offshore undersea cable focused primarily on servicing the country’s oil and gas industry.

This represents an increase of tens of terabits per second in planned and existing undersea fibre capacity around Africa in the last 12 months;  a very significant sign of optimism in the telecommunications market.

Terrestrial Backbone Fibre

Measuring the growth of terrestrial fibre networks in Africa is quite challenging as projects often have several phases of deployment, from financing to breaking ground to technical completion to inauguration, all of which may be deemed newsworthy.  A project can also have multiple construction phases each at varying stages of completion.  So what can we say about terrestrial fibre in 2015?  There were nearly double the number of announcements from 2014 with new projects from 19 African countries, comprising over 22 thousand kilometres of fibre and over 730M dollars of investment.  As you can see from the table below there are many gaps in information about network size and investment so these figures are likely extremely conservative.

Last year probably represents peak investment in national fibre backbones as virtually every country on the continent has embarked on national backbone projects.  Just as undersea cables have spurred the development of terrestrial fibre backbones, we can now see how this wave of investment is triggering investment in metropolitan fibre networks.

It is also possible to see that China dominates the terrestrial fibre network industry in sub-Saharan Africa.  Chinese investment has resulted in the development of tens of thousands of kilometres of fibre being built on the continent.  This is positive news but there are trade-offs that need to be considered.  One that concerns me in particular is that often fibre network investment goes ahead without due concern for how the network will be managed in the longer term and/or how it will be used to stimulate competition.  The World Bank, to their credit, place a strong emphasis on Open Access policies related to their network investments.  While fibre has spread rapidly across the continent, there is a large variance in its uptake.  Poor management or anti-competitive behaviour often limit the impact of this infrastructure.

CountryKilometresInvestment ($M)AnnouncementInvestmentContractor
Botswana10009.5Feb 2015??
Cameroon91640Oct 2015African Development Bank?
Republic of Congo52023Oct 2015World Bank (Central Africa Backbone)Huawei
Democratic Republic of CongoMar 2015World Bank (Central Africa Backbone)?
Ghana80038May 2015Danish International Development Agency (DANIDA)Alcatel-Lucent
Guinea4000238Jul 2015Eximbank (China)Huawei
Kenya1600Nov 2015Chinese Government?
Ivory Coast5000165Dec 2015??
Mali10646.5May 2015Maroc Telecom
Niger227593Jun 2015Chinese Government
NigeriaJan 2015
Senegal300085Oct 2015Huawei
Sierra Leone40015Apr 2015Exim Bank (China)Huawei
South Sudan1000Jan 2015World Bank
TanzaniaJun 2015China International Telecommunication Construction Corporation (CITCC)
Togo200Apr 2015Exim Bank (China)Huawei
Zambia5005Jan 2015
South Africa17.5Nov 2015
Malawi428Mar 2015World BankChina Telecom

Metro Fibre, Fibre To The Home (FTTH), and Video on Demand (VoD)

It is a perfectly natural development, as undersea then terrestrial fibre networks developed in sub-Saharan Africa, that metropolitan fibre networks would follow.  And the best news story this year on metro fibre has been the expansion of Google’s Project Link to a second African city: Accra, Ghana.  I’ve written here, here and here about why I think this is such an important project.  The Accra announcement signalled that Project Link is not just an experiment for Google but a strategy they plan to implement in other cities.

Municipally-driven metro fibre has moved more slowly elsewhere with few major cities engaging in the strategic development of municipal fibre networks but rather allowing individual operators build out networks.  There are a few exceptions.  South Africa has a number of metropolitan fibre networks including Cape Town, Johannesburg, Tshwane and Ethekwini which embrace a variety of different deployment models.  Cape Town has built out a municipal fibre network that leases excess capacity to the private sector.  In July of 2015, they released some impressive statistics as to how much money the new network has saved them.  Johannesburg opted for a Public Private Partnership (PPP) with Ericsson which did not work out well, resulting in the city cancelling their contract with Ericsson.  As of 2015, it is now a Municipally Owned Entity with plans for expansion.

The development of PPP-driven metropolitan fibre networks is apparently part of the fourth phase of Tanzania’s National ICT Broadband Backbone (NICTBB) plan but I have not been able to locate any additional detail.   In 2015, we also saw the spread of metro fibre outside of capital cities and into secondary cities such Liquid Telecom’s announcement of fibre investments in Kilifi, Kenya

Looking back it now seems inevitable that the development of metro fibre networks would lead to Fibre To The Home (FTTH) offerings but I don’t think anyone expected it to happen so fast.   Below you can see a list of FTTH announcements from 2015.  Two things stand out.  One, South Africa in particular is going through a massive wave of FTTH competition.  I listed a few announcements but it doesn’t really do justice to the explosive growth of competition in the last year from new FTTH speciality operators and existing telcos.  The second issue worth noting is the continued investment in FTTH from Liquid Telecom in a number of markets.  In August they launched Hai, an ISP offering FTTH service in Zambia with Rwanda and Kenya to follow shortly afterwards.

Ivory CoastAbidjanApr 2015Ivoire IP deploying FTTH for neighbourhoods via franchise model
KenyaNairobiMar 2015Liquid Telecom to Re-enter Home Market
RwandaKigaliOct 2015Rwanda: FTTH makes its way to Kigali
South AfricaCape TownFeb 2015Rich Cape Town suburbs join fibre-to-the-home race
South AfricaJohannesburgJun 2015Blairgowrie selects Vumatel for fibre
South Africa Cape TownSep 2015MTN home fibre coming to W Cape | TechCentral
South AfricaDurbanSep 2015MTN set to launch fibre-to-the-home in KwaZulu-Natal
ZambiaLusakaFeb 2015CEC Liquid Telecom launches Fibre to Home promising fastest broadband speeds in Zambia
ZimbabweVictoria FallsApr 2015 Liquid Telecom brings 100MBPS broadband to Victoria Falls
ZimbabweHarareMar 2015TelOne to connect 9 500 homes

And, of course, where there is FTTH there are Video on Demand (VoD) services.  New VoD services were launched in Zimbabwe but again the real stand-out was South Africa where Naspers, Vodacom, Telkom, and MTN all announcing services last year, seemingly in an attempt to get ahead of news that Netflix would soon be launching there.

Licensed Spectrum

The wireless world has been busy too, with operators across the continent rolling out new LTE networks.  Just as in 2014 we can see a mix of frequencies being used.   It roughly breaks down to a third of the deployments recycling 2G/3G spectrum mostly from the 1800MHz band, a third using 800MHz bands where previously CDMA networks dominated but which have now largely been abandoned, and a final third using 2.3 and 2.6GHz band some of which is new spectrum and some recycled WiMax licenses.

New for 2015 is the spread of spectrum auctions beyond Nigeria, which was for years the only country on the continent to carry out spectrum auctions.  Tanzania announced plans to carry out spectrum auctions in 2016 and Ghana successfully auctioned 800MHz spectrum.  While demand for spectrum is high in virtually every country in sub-Saharan Africa, regulators have been slow to make spectrum available.  This can be attributed to the challenges involved in carrying out a fair and equitable spectrum auction.  Poorly organised auctions can often result in either non-participation by key stakeholders or legal challenges later from those unhappy with the results.  In Kenya, the regulator chose to forgo an auction and assigned 800MHz to the largest operator, Safaricom.  This resulted in a challenge from competitor Airtel and Safaricom being asked to hand back some of the assigned spectrum to make available to other operators.   Morocco has also strategically avoided spectrum auctions.  In March, the Moroccan regulator, ANRT, made a massive assignment of spectrum to all the major operators making spectrum in the 800MHz, 1800MHz, and 2600MHz bands available to them.  This has the significant downside of not creating space for new market entrants but, on the other hand, it gets spectrum out there and a ‘use it or lose it’ or even a ‘use it or share it’ policy could be used as a safeguard.

There were many more LTE network announcements 2015 than 2014, more than double, bringing the technology firmly into the mainstream in Africa.

 GambiaNetpageApr 2015230040TDD
 KenyaSafaricomApr 201580020FDD
BeninSmile TelecomNov 2015???
 TanzaniaTTCLDec 2015???
 NigeriaMTNJul 201580020FDD
 Guinea BissauOrangeDec 2015???
 MauritiusMTMLMay 201518003FDD
 MalawiAccess CommunicationsApr 201580020FDD
 EthiopiaEthiotelecomMar 201518003FDD
 UgandaVodafoneFeb 2015260038TDD
 MoroccoMéditelJune 201580020FDD
 MoroccoMéditelJune 2015260038TDD
 TanzaniaTigoApr 201580020FDD
 CameroonMTNDec 2015???
 TanzaniaSmartSep 2015230040TDD
Nov 2015350042TDD
 BotswanaMascom and OrangeFeb 201518003FDD
 MoroccoMaroc TelecomJul 201580020FDD
 MoroccoMaroc TelecomJul 201518003FDD
GabonAirtelDec 201580020FDD
 South AfricaCell CSep 201521001FDD
 Beninbe.TelecomsDec 201518003FDD
 MoroccoInwiJune 201518003FDD
 MadagascarTelmaJun 2015???
TanzaniaSmart TelecomAug 2015230040TDD
MadagascarBlue LineJan 2015250041TDD

Unlicensed and Dynamic Spectrum

Perhaps the most profound impact of the spread of terrestrial fibre on the continent is the manner in which it is enabling alternative last mile access in an environment where the last mile has been controlled by those with access to licensed spectrum.  Unlicensed (WiFi) and dynamic spectrum (also known as television white spaces spectrum) are now coming into their own as terrestrial fibre networks offer the enabling backhaul infrastructure.

Of the two, dynamic spectrum has moved more slowly in 2015 as we wait for manufacturers to scale up manufacturing of dynamic spectrum chipsets.  There is an unfortunate Catch-22 where communication regulators are waiting for signs of large-scale manufacturing to move forward with dynamic spectrum regulation and manufacturers are waiting for clear signals from regulators before scaling up manufacturing.  I had hopes that South Africa or Malawi might have formalised dynamic spectrum regulation in 2015 but it seems that this will likely happen in 2016.

Progress in the form of new dynamic spectrum pilots took place in Mozambique and Morocco.  And in Kenya, the announcement of a potential 4 million dollar investment to expand the Mawingu television white spaces initiative was very exciting as well as the government of Mauritius’ announcement of a national wireless network based on television white spaces and WiFi technologies.

But these are just teasers of things to come as dynamic spectrum technologies mature and grow to scale.  The real story of 2015 was the spread of WiFi networks.  South Africa is the real stand-out in this case, with WiFi networks being announced in Cape Town, Durban, Johannesburg and Pretoria — and not just on the streets but on buses and in petrol stations.

Also exciting was the emergence of wholesale-only WiFi initiatives such as Project Link in Uganda, Naspers/DiData in South Africa, and BoFiNet in Botswana.  Given the limited spectrum available for WiFi this makes a great deal of sense in urban areas.

Finally, it is interesting to see the competing models for these WiFi deployments.  Project Isiswe believes that WiFi can only be delivered as a free service financed by government.  Startups like Argon Telecom in Kenya, which promises to deliver low-cost WiFi to Kibera, believe that there is a business model in very low-cost alternatives to mobile services.  Ultimately, governments, businesses, and consumers ALL derive benefits from connectedness and as a consequence there is truth in both propositions.

Country Technology Date News
Mauritius WiFi / TVWS Sep 2015 Maurice : le gouvernement veut utiliser les fréquences TV libres pour couvrir toute l’île en Wi-Fi - Agence Ecofin
Kenya WiFi Mar 2015 Kenya: Liquid Telecom invests USD 400,000 into Kilifi
South Africa WiFi Sep 2015 Total offers free Wi-Fi to customers in SA
Kenya WiFi Nov 2015 Kenyan counties eager to tap into Liquid
Kenya WiFi Sep 2015 Argon Telecom wants to halve the cost of data for residents of Kenya’s Kibera using low-cost Wi-Fi roll-out
Nigeria WiFi Jun 2015 Nigeria: Airtel rolls out Wi-Fi service in Lagos
South Africa WiFi Nov 2015 Free Wi-Fi for Gauteng’s East Rand
South Africa WiFi Oct 2015 City of Joburg launches municipal entity for free internet deployment
South Africa WiFi Jul 2015 Buses in South Africa get free Wi-Fi
Kenya WiFi Feb 2015 Liquid Telecom takes over Nakuru free Wi-Fi contract
South Africa WiFi Oct 2015 Cape Town free Wi-Fi inches forward
South Africa WiFi Nov 2015 DidataNaspers JV sets out Wi-Fi plans
South Africa WiFi Jul 2015 Free Wi-Fi for Cape Town buses
South Africa WiFi Oct 2015 Tshwane free Wi-Fi reaches a million users
Zimbabwe WiFiApr 2015TelOne rolls out Metro Wifi
Kenya TVWS Jul 2015 Kenya’s TV white spaces network to be expanded
Morocco TVWS Nov 2015 La technologie ?TV White Space? pour connecter les ?tablissements scolaires en milieu rural au Maroc - Tic Maroc
Mozambique TVWS Aug 2015 INCM, Internet Solutions to collaborate on ‘white space’ pilot
UgandaWiFiDec 2015Bringing Better WiFi to Kampala
AlgeriaWiFiApr 2015Algérie : la wilaya de Bechar au centre d’un projet d’accès généralisé au wi-fi public
BotswanaWiFiMar 2015BoFiNet Rolls Out 600 WiFi Hotspots
SeychellesWiFiDec 2015Internet provider proposes to make Seychelles a Free Wi-Fi Nation

Mixing Things Up

Another interesting trend in African telecoms in 2015 has been the increased entry of non-telco players into the telecommunications infrastructure market.  Internet companies like Google expanding Project Link from Kampala to Accra and perhaps beyond is a great example of this.  Elsewhere, other interesting things were happening, such as the announcement that tower operator, Helios, would acquire Telkom Kenya.  If tower companies start thinking about leveraging their assets, who knows what might happen.   Power companies, too, are flexing their muscles.  Kenya power announced it would begin offering retail services and in Botswana the power utility announced it was considering opening up its network.  Taken in sum this looks like very good news as markets that have traditionally been the province of just a few large players are now exposed to competition from a variety of directions.


What can we take away from the above?  For me it is clear we are just beginning to see the profound the impact of fibre optic infrastructure on the continent.   Mobile infrastructure in Africa took ten years (roughly 1994-2004) before its real impact begin to be felt.  I believe we are now starting to feel the real impact of fibre infrastructure that began with the landing of the Seacom cable in 2009.  It is bringing true high-speed, affordable infrastructure to the continent and enabling both mobile and alternative next-generation access networks.  Do you agree? Disagree?  Please let me know in the comments.  Also, as with last year,  please help me out with news I’ve missed or issues I have misconstrued.

This work would not have been possible without the support of the Network Startup Resource Center

Post publication edits:
Addition of BoFiNet wholesale WiFi deployment and Seychelles WiFi initiative