⇓ More from ICTworks

If You Really Want to Scale ICT4D to Millions, Abandon Africa

By Wayan Vota on August 7, 2015


We all talk a good game about how we want to scale ICT4D programs. How we want to reach 100,000 or even +1 million constituents with our cool technology solutions. If that’s the true goal, and we’re honest about reaching it, we really should abandon Africa for India or China.

Go to China or India

The key to scale is people of course. And most surveys put China and India at the top of population counts with over a billion people each. Africa, as a continent, also has over a billion people. Africa, however, is not a country, so any solution needs to be reflective of 50+ countries and administrative regions. China and India have their own political barriers too of course. However, Uttar Pradesh, just one state in India, has 25 million more people than Nigeria, sub-Sahara’s most populous country.

Now let’s say you could cross political barriers without too much hassle, working in Africa, you would still have to surmount steep language barriers. Nominally, most Africans speak English, French, Portuguese, or Arabic, but in reality, those are languages of the elite. You would need to localize to hundreds of tribal languages to reach millions of Africans, or just English and Hindi to reach 540 million Indians, or Traditional Chinese to reach almost all 1.3 billion mainland Chinese.

Or Redefine Scale

Finally, let’s really be honest about “scale”. This term doesn’t mean we need to reach millions, go continental, or even national. Reaching scale can mean a majority of community members in a single village, town, or city. Or just your key constituents, no matter their quantity.

When we understand that scale is relative, we can let go of the need to abandon Africa to reach millions, and accept that scale is best defined on an individual basis, one changed mind at a time.

Filed Under: Thought Leadership
More About: , , , , , , , , ,

Written by
Wayan Vota co-founded ICTworks. He also co-founded Technology Salon, MERL Tech, ICTforAg, ICT4Djobs, ICT4Drinks, JadedAid, Kurante, OLPC News and a few other things. Opinions expressed here are his own and do not reflect the position of his employer, any of its entities, or any ICTWorks sponsor.
Stay Current with ICTworksGet Regular Updates via Email

20 Comments to “If You Really Want to Scale ICT4D to Millions, Abandon Africa”

  1. Hello Wayan, I think you have a point but i do not agree with you all your comments. There is a way you handle those barriers you mentioned in your article. Just remember that some of the Africa countries are organized in Regional groupings and by working with those entities you can scale up best practices across the region and most countries under the regional groups or communities often do that.

    Its an open discussion and i think we can work out things. Check our Millennium Villages approach and CHWs success.


    • Wayan Vota says:

      Agreed – COMESA, ECOWAS, and many cultural and linguistic commonalities do exist and are helpful. And yes, scale can be achieved within and across African countries. Better yet, would be an industry-wide acceptance that “scale” is a relative term.

  2. Is this a bait article (good headline – no substance), that you wrote because you have to churn out something on a friday, apart from “learning” that Uttar pradesh has a big population, is there anything else to learn.

    • Wayan Vota says:

      Cavin, I’m glad my “click bait” headline and provocative text moved you to comment. Thank you.

      But this post wasn’t something I just threw together. It is a direct reaction to the many. many times I’ve heard good ideas shot down because they were not going to reach some preconceived notion of scale, usually a 10-100x factor larger than what’s possible to be effective.

      My point, maybe not made clearly enough, is that we do great harm by thinking everything has to reach millions of users to be successful.

      If you want a deeper analysis of how scale and other development jargon is harming us, here is a great post by Patrick Fine, CEO of FHI 360: Words Count!

    • Internet Comment says:

      Agreed, this is even worse than a buzzfeed listicle.

  3. Charles Marc says:

    Hello Wayan. I am an ardent reader of your posts and ICT works in general but on this occasion i am inclined to disagree with you on a number of things in this article but most importantly the reasons you give to abandon Africa. I know you mean well-Yes we are based on many tribes and regions but many ICT4D programes have scaled in these regions and these models have been replicated in other regions for example the Grameen CKW (Community knowledge worker) in Uganda taken to Kenya and Columbia. Like Tiamiyou RADJI says in every community there are barriers and we can innovate around these barriers and ultimately scale these best practices based on the learnings from those areas you suggest we abandon-Don’t give up on us

    • Wayan Vota says:

      Charles, in no way am I actually advocating to give up on the people of Africa. I am angry that we too often claim something must scale to millions to be effective. If that’s the real goal – millions of users – then yes, let’s work where its easier to reach that number.

      But if we want to have real impact – which is best measured per person – then let’s accept success can be found in one African community at a time.

  4. Lauren says:

    Great post! I am glad someone has finally pointed out the obvious truth about “scale.” About time we cut through development jargon.

    • Joyce says:

      If development people spent less time on development jargon and academic concepts like scale, maybe there wouldn’t be so much waste. I believe this article is spot on, although I think the message was somewhat buried.

  5. Dipanjan says:

    It looks like you are asking people to build a solution first and look for problematic places where the solution would fit. Whereas the better approach is to identify a problem in a place and then design a solution for it. Just because Africa does not have the numbers does not mean its problems need not be solved!
    The second part of the post tries to counter it but it is a little too late and subdued IMO.

    • Wayan Vota says:

      Dipanjan, that’s the issue: we often find great problems to solve but are told that we are not reaching enough people, so we try to scale it beyond those it’s targeted for, and then are told we failed because we didn’t scale.

      • Dipanjan says:

        Hi Wayan,

        Thanks for clarifying. But this somehow didn’t come out in the post. First time readers might be misled. I follow your blog and know where you are coming from.

  6. I wonder if ICT4D (in my specific case, Digital Health) is already, to some extent, “at scale.” It’s just fragmented. If you were to add up all the numbers from individual projects they would be impressive. And to follow that line of thinking a bit, the key to reducing that fragmentation might then be to work towards standards and interoperability. But this is just an ill-formed thought I had yesterday while being challenged, like we all have been (repeatedly), on the issue of scale…

  7. Ed Gragert says:

    Hi Wayan,

    Thanks for reminding us that “scaling” means going to a higher level than what is being done now. All too often, policy makers and donors expect a national or regional roll-out of a successful pilot project. But, a doubling, tripling, ten-fold, hundred-fold, or even multi-community expansion are also viable examples of scaling up. Expectations need to be in line with realities. And with this in mind, no part of the world can be abandoned.


  8. Mireille Nsimire says:

    Hi Wayan,
    Thank you cooling down your Article by redefining the Scale. From our on point of view I agree with on the fact that the “scale”. doesn’t mean we need to reach millions, go continental, or even national. Reaching scale can mean a majority of community members in a single village, town, or city. Or just your key constituents, no matter their quantity. And that it our daily motivation to concentrate our ICT4D Work for Africa.
    As An African I trust that we shall get there.

  9. I think the barrier you gave is language. When it comes to language, the first barrier is between technologist and development practitioners. Those trying to address this through making development practitioners ICT4D competent have achieved limited success because they use computer science language or teach an application (i.e. mHealth, mAgric etc.) rather than provide a generic ICT4D tool that works across a broad front of ICT4D applications.. What we are working on is to develop a Participatory Rural Appraisal (PRA) based framework as a standard User-Centered Design tool for ICT4D. The importance of this is that the framework is built from the ground using a tool that is familiar to development practitioners and has been validated as useful in crossing language barriers between cultures and economic classes (i.e. PRA has scaled). This, then cancels out language barrier as the reason to abandon Africa.

    • Wayan Vota says:

      Language is certainly one barrier, as is culture, etc. Also there are political boundaries that act as barriers – at least 50+ on the continent. You also point out the barriers in our minds as development practitioners – people who should really be on the same team.

      • Mireille Nsimire says:

        I don’t think that the language is a barrier in Africa If You Really Want to Scale ICT4D. for those who don’t know Africa is organize based on community ( we have north, South, East, West and Central Africa) and those community speak almost the same language.
        currently the National ICT4D strategy is being develop to allow our leaders to facilitate the growth of ICT sector in different African countries.
        some African countries have already established they national ICT4D strategy.
        In most cases the political boundaries are based on the ignorance of our leaders concerning ICT4D concept. we have started to involve them in several ICT4D activities across our countries, and regions hoping to provide to them a minimum understanding of what we are committed to.

  10. If a project is not able to scale, you should address why it’s unable to scale instead of blaming the weather, population size, language etc. Mobile money has been a success in Kenya and other African countries, but failed to take off in Ghana, due to the populace preferring hard cash which they can touch. It’s not an issue of small population or tribe or language. It’s an escapist mentality to blame other things instead of taking a critical look at yourself. When you point a finger at someone, its very easy to forget that it’s only 1 finger pointing in his/her direction, the other four fingers point in your own direction.

  11. Don Osborn says:

    Roundabout way to call into question the focus on scaling, but interesting (although I wouldn’t use the term “tribal languages”).

    From my experience as a non-African who has worked in and on Africa for many years, it does seem that more “artisanal” than “industrial” approaches to development (including ICT4D) may be appropriate much or most of the time. Take reforestation in the Sahel, for example – foreign funded projects for years focused on “fast-growing” foreign species in large scale projects – scaled up, if you will – while neglecting sylviculture of all but a handful of native species, with mixed overall success, and later recognition that (1) some of those foreign species pose long-term problems (eucalyptus, prosopis), and (2) some useful native species are harder to find. A local-first and more balanced approach might not have had as dramatic an initial result, but the longer term outcome might be better.

    Likewise, non- or poorly-localized ICT4D might produce numbers, but not results. Maybe long-tail should be the paradigm, to “scale” multi-locally?

    Long-tail is also relevant to understanding the linguistic realities for decisions about localization. At no matter what level one looks (region, country, province, district, village), the population of speakers of languages (1st, 2nd, or both) is always an asymptotic distribution. But the languages at the head of the distribution might be different at a local than national level, or from one province to another, etc. Researching a potential project area in this way might produce manageable numbers of localizations necessary to serve most of the communities (rather than being discouraged by reading about hundreds of African languages).