We attended the ACM Dev conference this year, which by good fortune was happening in London! We’ll be putting up our notes shortly1, but this blog post will discuss some of the higher-level considerations that came up as a result of the conference.

A hammer seeking nails

The first day of the conference was the GAIA (Global Access to the Internet for All) workshop. There was a lot of good technical work at the workshop, and it was encouraging to see empirical evaluations of the effectiveness of the projects. However there was little discussion about whether the projects actually improved welfare, or indeed whether they had an effect sustainably or counterfactually.

This is probably to be expected from a workshop that is specifically about extending internet access to all, but it was the first occurrence of a recurring theme: as technologists, we are in danger of being the person with a hammer who is looking for nails, without considering whether that is what is currently most in need of attention.

We consider this to be a serious problem. Why should we focus on problems that are addressable by technology (or, more narrowly, software)? Shouldn’t we look for the most pressing problems in any field, and then consider what we can do about those? We’re still thinking about what the implications are for our strategy, and we’ll discuss those in a later post.

Start with the needs

When discussing the project, a common observation was that it can be very difficult to accurately discern the needs of those on the ground. Arguably, there is no substitute for actually living in Africa to become acquainted with the needs of people who live there. Ken Banks, in particular, was even critical of directed surveys - people will tell you what they think you want to hear, and so in order to really find the problems you have to actually observe them yourself.

This is wise advice whenever you are starting a project - many successful businesses are based on the needs that their founders saw, personally. And so we would certainly agree that it’s important to start from the needs of the people you will actually be helping (rather than starting with your “hammer”). But such knowledge may be transferrable. There are many people working in the field, and it’s unlikely that they are acting upon all (or even the most pressing) of the problems they have observed. We should still expect to be able to generate fruitful communication (and collaboration) between those people and people who would be interested in implementing their suggestions.

Furthermore, by aggregating this information we can start to evaluate and prioritize possible courses of action. While being on the ground may be very helpful for spotting a problem, it’s unlikely to lead you directly to the most important problem. To discover that, we need to bring together the work and observations of many people.

Be realistic about the environment

Africa has astonishing mobile phone penetration - almost 70% of Africans have a mobile phone. But far fewer of them, only 15%, are smartphones, and mobile internet is very expensive. So for an application to have a broad reach, even if it is built on mobile technology it must be built on the lowest common denominator. That means voice or SMS (or maybe USSD), and so while an internet-based application may be appealing, it may not be realistic.

Other logistical challenges are simply unfamiliar to most Westerners. Africa has very low population densities compared to most of the developed world, and even a simple installation process can be a huge problem if it is a two-day drive to the installation site.

Politics, as well, can prove a significant barrier to implementation. Many developing countries have government-sponsored monopolies in certain areas, and any attempt to undercut them may be viewed quite unfavourably. In general, technology initiatives that empower the poor are likely to upset at least some social power structures, and failure to consider the wider social consequences of an intervention can doom it to eventual failure.

Once again, the problem here is knowledge. All of this information exists already, but it is not accessible to the right people. We should expect to need this information to succeed, but we should also be able to acquire it by talking to people who already know it.

Sustainability through business

The ICT4D community is very concerned with sustainability. Many of their projects run with limited academic funding, and so amount to pilot projects that can fizzle out after their term is complete. This is criticized as a failure of sustainability - if the project only runs for a short time, then it has not changed anything in the long run.

Pilot programs are certainly vital, especially in research. It may be necessary to investigate and evaluate many options before the best solution is found. However, once we’ve hit on a good solution, we want to implement it with a long-term sustainable solution.

The best option seems to be to turn the program into an actual business. Businesses are inherently sustainable as long as they are able to sustain their profitability (and if they can’t, perhaps the need for the program has passed). Furthermore, they will naturally scale up successful interventions without the need for additional donor funding (or if they do need funding, this can be acquired through less constrained channels).

Some of the people we spoke to do work on spinning off their projects into sustainable businesses, usually in collaboration with local partners. This could be an opportunity for Western entrepreneurs to make themselves useful, either by assisting existing spinoffs, or by taking over directly from the researchers.

Innovation in business models could also be very valuable by allowing more programs to be feasible as businesses. Programs which have a high social impact but are difficult to run as a for-profit business can only be implemented by non-profits, with all the sustainability problems that that entails. If we can expand the sphere of potential businesses we open up more opportunities.

Why should we expect to be better than the poor themselves?

It’s well known that people in developing countries are often astonishingly entrepreneurial. They do very creative things with the resources available to them, and often find niches for profitable businesses in remarkable places.

Furthermore, access to the internet and computing resources are also spreading. That all adds up to a climate in which we might expect African entrepreneurs to fill many of the needs that might be filled by a Western technologist. So we should have good reasons to believe that we can supply a need better than the local market, whether would do that with a novel business or a non-market solution. Why do we think we can do better than the people on the ground?

This is a pertinent question, and in many cases if a niche hasn’t been filled we should take that as a tip that perhaps it’s not as good an opportunity as we thought, or that there is some obstacle which is currently invisible to us.

However, there are still things which Westerners have to offer. For one thing, we have access to resources that the poor still do not, whether that be capital, markets, greater pools of talent, knowledge, or just coordination. There may well be high-impact opportunities in spreading access to these tools (for example, running programming training courses), thus enhancing the ability of locals to solve their own problems.

Secondly, many opportunities may have mixed benefits, in that while they are profitable, they also have a large amount of positive spillover. The lack of profitability can be because it is difficult to capture the value of the program, perhaps because the beneficiaries are very distributed (as is common with infrastructure projects), or because the value accrues in the longer-term future (as may be the case with education programs). We should expect such market failures to be relatively undersupplied relative to opportunities that are more profitable. But from the point of view of the altruistic entrepreneur, these are ideal opportunities, because positive externalities are also desirable.


It’s clear that there’s a lot we can learn from the ICT4D sector. We’re optimistic that we’ll find some useful opportunities here, and we’ll be writing more about this area, as well as the implications for our strategy going forwards.

  1. We intend to make all of our notes and materials publically available, even if they’re provisional or later revised.