Theory of Change: The Village Telco

Having funded and watched and occasionally participated in the wireless hacker space in Africa for the last few years, I have the sense of “waiting for the next leap forward”. Wireless hackers have been successfully building cantennas, woktennas, and waterbottletennas, to name a few. They’ve been flashing Linksys Routers with a variety of free firmware replacements and in general tinkering with WiFi in a very cool way. I say cool because I think tinkering is one of the best ways of learning about something and also because taking something apart is a way of demystifying technology and ultimately making it something that serves you as opposed to the other way around.

There have been quite a number of initiatives in this space, some grassroots, some donor initiated and most have been successful in crossing that first bridge, of developing a level of comfort in adapting a commodity technology to serve a local purpose. As I say, I think this alone is an important outcome. Having said that, it feels as if a certain “so-what” stage has been reached. If you have good connectivity, WiFi can be a natural way to distribute access and to forge local broadband links up to x kilometres away. However, connectivity in Africa is still both scarce and expensive. This means that there is limited utility to basic wireless solutions. Something more generically useful and applicable is needed.

Think Fon but for Rural Voice Services in Africa

That thing, I think, is a “village telco” and by that I mean an easy-to-use connectivity solution that provides extensible local telephony with the possibility of upstream voice and/or data connectivity via POTS, mobile, or other IP services.

In a perfect world, this would be an out-of-the-box solution which provides an Asterisk PABX/telco along with a number of preprogrammed handsets which would offer instant local phone calls and voicemail to anyone within the radius of a WiFi antenna. This is not a Wireless ISP in a Box (WISP) with the possibility of some VOIP services overlayed on it. This is a local telephony solution with the possibility of upstream network connectivity for both voice and data. I make this distinction because voice is still the killer app i.e. delivers the most general utility and I think you need to deliver this service before all others.

What I am imagining is something like the Fon initiative but targeted at providing low-cost local voice services in Africa. Data is cool too but voice is the minimum entry point. Instead of buying a Fon wireless router, you as a local entrepreneur would buy a small Asterisk PABX that comes with a dozen or two pre-programmed phones. The whole thing should work out of the box with people simply needing to associate themselves with a phone extension. Voila! instant local phone calls and voicemail. Part of my theory of change is that a local phone service on its own would provide enough utility to warrant the investment. In December, I was at the Global Knowledge 3 conference in Malaysia where Ross O’Brian of the Economist Intelligence Unit mentioned statistics from Afghanistan that the majority of phone calls were local to the community they were initiated from. That started me wondering what percentage of phone calls are local in Africa, in urban, peri-urban, and rural areas. If anyone has any evidence one way or the other I would love to hear about it.

Having said that, delivery of local voice and voicemail is just the tip of the iceberg. Ideally, a “village telco” device would easily connect to a regular phone line, mobile service, or Internet connection to offer upstream voice and/or interconnectivity. It would also offer a simple billing system that could be adapted to a variety of billing models based on pay-as-you-go, monthly accounts, or even barter.

This is NOT the Nokia-Siemens Village Connection

While I applaud the ingenuity that has gone into designing the Village Connection and the spirit of entrepreneurship that it hopes to engender, it seems to me that Nokia-Siemens’ decision to distribute the Village Connection through the mobile operator supply chain means that the Village Connection is only going to help entrench market control for mobile operators.

A “village telco” by contrast would allow villages/communities/organisations to choose their own interconnections and upstream providers, to use VOIP where it is most cost-effective and mobile/POTS services. It would also allow “village telcos” to interconnect among themselves.

It is not just in the realm of consumer choice that a “village telco” would be different in a very important way from the Village Connection. Based on a shared, transparent Open Source/Open Hardware model, the “village telco” would allow widespread entrepreneurship and grassroots innovation. Necessarily, a “village telco” would need to easily evolve to adapt to local realities. The Open Source model for the “village telco” creates a trusted open space for innovation to be fed back into the design.

This is not to say that that the “village telco” idea is not commerce-friendly. There are plenty of opportunities to provide upstream services to a “village telco”, to interconnect, to offer value-added services, etc.

Coexistence with Mobile Operators

Because a “village telco” would offer non-mobile (or at least limited to local WiFi coverage) services, it should be able to co-exist with mobile providers. There would still be a natural aspiration path to having a personal mobile phone that offers more flexibility (and perhaps status). Once connected to a mobile operator, every customer of the “village telco” has the potential to generate revenue for the mobile operator by dialing phone on their network.

But Why Stop at Voice and Data?

Pundits have been touting financial transactions as the next big thing for mobile markets and indeed the early success of initiatives such as mPesa appear validate these predictions. A plausible next step for a “village telco” would be to support a transaction system that could serve as a financial/barter proxy for a village/community. It open the possibility of providing local solutions to local economies. But perhaps I am getting ahead of myself.

African Undersea Cables

Mostly in order to keep them all straight in my head, I have compiled a list of African undersea cable initiatives and their features, investors, etc.

Cost (USD)
650 million 265 million 82 million 2 billion? 865 million
Length 13,700km 10,000km 4,500 km 13,000 km 14,000 km
1.28 terabits/s 640 gigabit/s 40 gigabit/s 3.84 terabits/s 2.5 terabits / s?
June 2009 2009 April 2009 2010 Q4 2010
USA 25%
SA 50%
Kenya 25%
Operators 90%
Kenya Gov 20%
Private Investors 65%
UAE Eitelsat 15%
Infraco 26%
Private Investors 74%
US 100%?
Items in red have been updated. Last update: 9 May 2008

Investor detail:

Seacom (

Industrial Promotion Services (25%), an arm of the Aga Khan Fund for Economic Development (USD 75 million)
(Kenya – founded by Prince Karim Aga Khan IV of Pakistan)

VenFin Limited (25%) – USD 75 million)

Herakles Telecom LLC (backed by Blackstone) (25%), New York-based lead company, no website (USD 75 million)

Convergence Partners (12,5%) – USD 37.5 million

Shanduka Group (12.5%) – USD 37.5 million


EASSy is 90% African owned although that ownership is underwritten by a substantial investment by Development Financial Institutions (DFIs) including World Bank/IFC, EIB, AfDB, AFD, and DfW. Total DFI investment is apparently $70.7 million, with $18.2 million coming from IFC, 14.5 million from AfDB. This is a smaller amount than the originally advertised $120 million investment from DFIs.

South African investors in EASSY include Telkom South Africa ($18.9 million) , Neotel, and MTN.

There are 26 telco operators in total invested in EASSy.

An SPV created to facilitate. open access will be the biggest shareholder, with 46%.

In Jan 2008, VSNL announced an investment in EASSy


Currently, 85 per cent of the project is owned by the Government of Kenya and the rest by Etisalaat of the United Arab Emirates. The Gov’t of Kenya is looking for local investment in the cable. According to Business Day, Safaricom is a likely shareholder at 30 per cent with the government retaining a 20 per cent stake. The rest is up for investment by regional operators.

Uhurunet / Infraco

African telecom operators (45%), Nepad (30%), International investors (25%)

Confirmed and rumored investors:
• 5P Holdings (US)
• International Development Office of the government of Ras Al Khaimah (an emirate in the UAE)
• Pan-African Infrastructure Development Fund (PAIDF)
• Phelps Stokes Fund (American 501 (c)(3) nonprofit
• Sheikh Saoud of Qatar (rumored)
• South African ownership: Telkom, Neotel and MTN together own 27%


A US company founded by African and African-American group of telecom veterans. Financial backing from West African bank and

Ubuntu and Descartes

When I first encountered the Southern African word ubuntu a few years ago, it instantly resonated with me. A person is only a person through other people…. the suffering of one person diminishes all of us. Or as Archbishop Desmond Tutu puts it, “a solitary human being is a contradiction in terms”. A notion that the English language doesn’t seem to have quite the right colours to paint. It takes a lot of mixing of English words to express that one African word. It resonated with me as a human being as an acknowledgment of our connectedness of our human responsibility to each other, even to all living things.

More than the emotional connection to the word though, it made me think about the philosophy I studied at university. It reminded me that musch of the western canon of philosophy was based on the principle that the indivisible unit of humanity was one human being and that everything one needed to know could therefore be deduced through a process of introspection. Descartes’ famous expression “I think therefore I am” is the classic expression of this principle.

But why should that be true? Why should humanity be discernible from a single human being, any more than wetness can be discerned from an individual molecule of H2O. Of course we are more complex but why should humanity be an individual characteristic as opposed to a collective one. Something that only emerges out of human interaction.

My introduction to the concept of Ubuntu made me think that Descartes took a big philosophical left turn a few hundred years ago that western philosophy has yet to recover from.

It was both inspiring and validating to hear John Seely-Brown express the same notion in an Edutech podcast with Steve Hargadon. In it he says,

“…my own thinking for now 30 years is kind of the shift from a Cartesian point of view of I think, therefore I am where knowledge is a kind of substance that gets kind of poured into your head to build up stocks of knowledge in your head..supposedly, versus we participate, therefore we are in that in participation with others that we come into being and that could be psychoanalytically true but it is also in participation with others that we start to internalise our own understandings of the world and learn…”

Right on.

Funding Projects – Group Consensus Facilitation

The Context

Acacia team

The IDRC Acacia team in action

Every year in our team at work, we get together to discuss our research funding budget. Each team at IDRC is responsible for producing a yearly “pipeline” of projected research projects to fund. Our team is pretty spread out, in four offices across the African continent and in Ottawa. We rely on a face-to-face meeting every year around March to talk through everyone’s proposed projects and come up with a coherent set of projects that are in line with our 5-year strategic plan. The trouble is that these discussions, even with the best of intentions, often end up being a bit of a bun-fight with people defending their own projects and being critical, often due to lack of time to discuss the projects in sufficient detail. While we have improved incrementally in the way that we approach this, this year I hoped we could make a leap forward in how we approached this. My goals were the following:

  1. Have each team member thinking about the big picture, about the body of knowledge that our collective funding is going to contribute to and not just their own projects.
  2. Find a way to deepen the understanding of team members of each others’ projects in an efficient way.
  3. Not waste time arguing minutia of projects where there is already broad agreement i.e. focus our face-to-face time on the most contentious proposals.

I should mention that we did have some virtual team (text) chats via Skype to review the proposed projects but we limited those sessions to asking questions for clarification as the synchronous online text environment is not a great environment for debating complex issues. This gave everyone at least a nodding acquaintance with the overall project pipeline and gave team members clues as to areas that would need further elaboration prior to the face-to-face team meeting. I nabbed the idea of printing each project on a single sheet of paper by accidentally walking in on the team meeting a month ago.

As a setup for the meeting process, we had the title and description of each proposed project printed in as large a font as possible on a single sheet of paper. These were stuck up on a wall of our meeting room, in no particular order, spread out as much as possible. Each project description could then be read from a few feet away.

So, on to the day itself. After an Open Circle (OpenSpace style) we moved on to the first session of the day.

1. Articulating the Research Questions

Articulating the research questions

Khaled and Edith working on research questions

This session was pretty straightforward. Team members broke out into pairs and interviewed each other to articulate the key research question of each of their proposed projects.

In the past we have found that focusing a discussion around the project research question has the effect of sharpening the focus and understanding of the project in general. In this case, it also helped team members to better understand each others’ projects, albeit a subsection.

We spent about an hour doing this during which each team member wrote out the research question for each project on a sheet of paper and stuck it up on the wall under the relevant project. To the left you can see the wall full of project descriptions with the research questions written and posted underneath.

2. Social Tagging

More tagging...

Mike Jensen tagging projects

Now we come to my favourite part of the day. I am an avid user and a strong believer in the power of social bookmarking as well as the power of emergent taxonomies or folksonomies. I wondered whether it was possible to recreate some of the power of online tagging into a face-to-face environment. Thanks here to Nancy White, Alison Hewlitt, and Beth Kanter for pointing previous experiments that Beth had done on “live tagging”.

We adopted a slightly more prosaic approach than Beth. Each team member got a pad of Post-It notes on which to write their “tags” and they were instructed to post them around the projects. This was a process that had multiple benefits:

  • On a very basic level it forced team members to read project descriptions and research questions (again or in some cases for the first time). It deepened (in a fun way) the understanding of each project without engaging in debate.
  • It encouraged team members to think abut projects on a meta level thinking about themes and key issues
  • It allowed common threads to emerge from the tagging process.

This part of the process was a real success. It achieved all of the above and more. The biggest challenge in this session was to communicate clearly at the beginning what sorts of things one might write on a tag. For those who hadn’t used before, it was not immediately obvious what a tag represents. The best way I could think to describe tags was as a kind of aide memoire. I asked team members to think about descriptive terms they would use about the project that would help them find it again if they were looking for it. I highlighted the fact that these terms might well be different for each person. I think I still need to work on a succinct explanation of tagging as an introduction.

The tagging exercise took us through to lunch time. At this point I felt like we had achieved at least a couple of our goals, namely to deepen understanding of the projects and to get the team to begin connecting the dots and thinking big picture.

Stoplight Exercise

Stoplight exercise

3. The Stoplight Exercise
The purpose of this exercise was to begin to highlight projects where there was broad agreement and more importantly where there was very little agreement. Everyone was given a page of colour coding labels and was asked to mark projects with a green, yellow, or red dot. Green for “I am happy to support this project”, Yellow for “I have some questions I’d like answered first”, and Red for “I have issues with this project proposal”.

This exercise was intended primarily to identify projects where there was a lack of consensus among the team. The intention that this would allow us to focus the majority of our efforts on fundamental areas of disagreement and not let time get away from us by spending too much time discussing smaller issues simply because they came up first. In a way the stoplight exercise was a kind of coarse filter for the final “Gradients of Agreement”.

4. Gradients of Agreement

Gradients of Agreement

Chairs as gradient markers

This exercise was a vast improvement on any previous methods we have used to seek consensus. At a suggestion from Alison, we drew on Sam Kaner’s “Gradients of Agreement” from his excellent book “The Facilitator’s Guide to Participatory Decision Making”. Kaner suggests that disagreement is seldom black and white; that there are gradients of agreement. He offers 7 levels of agreement ranging from full-on endorsement to outright blocking. The interesting parts were the middle gradients which expressed sentiments such as “disagree but unwilling to hold the majority back” or “I can live with it”.

Each gradient was printed on a single sheet of paper. Originally I had thought to tape the sheets to the floor in a line in front of the projects on the wall but it turned out to be better to stick them to chairs and line the chairs up. See photo at left.

We then initiated a discussion on proposed projects, starting with those projects that had attracted the most red dots and working backwards from that. We began each project by asking team members to stand by the gradient that most closely reflected their feelings about the proposed project. While team members were asked to move dynamically from gradient to gradient, we found it was necessary to stop from time to time specifically to ask people to reflect on their “gradient”. Kaner goes into detail about the process and I can’t recommend the book enough. We carrying on discussing each project until it was evident their was sufficient agreement or it seemed likely that no consensus was likely. In general we assumed that each participant had veto powers but no one ever felt compelled to use them.

Improvements for next time
There is still lots of learning to be done in implementing the above exercises. I think each of them were worthwhile and each built well on the previous exercises. They added up to a very productive day and a stronger sense of vision and of common threads across our programming.

One area that we didn’t account for was how to deal with projects that are very similar. I think a separate clustering process linked to the tagging exercise might be a useful addition. We would have to change the way that we physically represent things on the wall though as it would have been too much work to move all the pages and tags around. Also, I would want to think about what to look for in clusters. Maybe just looking for areas of substantial overlap.

It is tempting to think of moving this whole process online. It think it would work very well and might facilitate virtual decision/consensus making processes. 🙂 Perhaps when I have some spare time…

All comments and suggestions welcome.