Where Africa and Technology Collide!

Building Mobile Apps for Africa

(Note: I’m going to ask you to contribute to this at the end, so start thinking…!)

I just got back from another trip to Kenya. This time I spent a week with Ken Banks as part of a workshop on social media in Africa, put on by some really smart people (thanks Mika and Linda!). As generally happens when we get in a room together, we start talking over some ideas that are bigger than what either of us are up to at the moment.

This time the discussion revolved around developing mobile phone applications in, and for, places like Africa. It’s becoming quite popular to create mobile products and services, but it’s still fairly new. It has only been a couple years since we first started seeing applications focused on this specific kind market. What have we learned? Are there any best practices on design and implementation? Is there a notebook for new developers to go to to learn what to do (and more important, what not to do)?

Anthony at SM4SC

19 tips from someone who’s been there

As Ken states, “In my experience, many social mobile projects fail in the early stages. Lack of basic reality-checking and a tendency to make major assumptions are lead culprits, yet they are relatively easy to avoid.

If there’s anyone who knows this field it’s Ken. He’s not only a thinker in this space, but he’s a doer as well. His application, FrontlineSMS, has cut it’s teeth here and he’s had to answer all the hard questions, which everyone else has benefited from. He has successfully created a basic platform that many other applications can build on.

Make sure you read Kens observations and tips on building mobile phones for developing countries.

Here are a couple:

  • Never let a lack of money stop you. If considerable amounts of funding are required to even get a prototype together, then that’s telling you something – your solution is probably overly complex.
  • Ensure that the application can work on the most readily and widely available hardware and network infrastructure.
  • Bear in mind that social mobile solutions need to be affordable, ideally free.

My contribution

  1. The next generation of Africans are more mobile literate than you (or me), so when you develop something make sure you keep it open enough for them to evolve its use.
  2. Develop for the common denominator – that is SMS services only. If you have the time, and see a need later, then go for the fancy Java apps.
  3. Data services, like SMS are a good starting point, but don’t overlook the use and integration of voice. This is especially relevant in areas where local language dialects and literacy are an issue.
  4. If you can, provide a basic service, and let the local users develop a plan for how to use it in their area.

Your contribution

Here’s where you chime in on what you think people should know before they build a mobile phone service or product for Africa. Got any tips for? Lessons to remember? Make sure you do/don’t do something?


  1. Thanks for picking up on this, Erik (as you always do). Looking forward to seeing how this goes – I think there’s a lot of potential to put together a paper (or more) on this if we can aggregate experiences.


  2. A great read, both yours and Kiwanja’s. My take is that developers would be prudent to begin considering building web-apps as well, for the mobile phone.
    I think that the projected massive bandwidth upgrade via fiber, at least in East and Southern Africa region, with ETAs in the 8-15month time frame, will be utilized primarily by mobile phones. PC penetration has been outpaced by mobile and is growing rather slowly and I think will be tethered mostly to office users and technologists that use them for work.
    A niche market? Perhaps at the onset, but I think such development will be forward thinking with the next 4 or so years in mind. And boy, does time fly quickly…

  3. still pursuing my dream app (traffic) , i mentiond this year so the tips much appreciated

  4. @bankelele – I’ve got someone to introduce you to who is interested in exploring the app you had on your mind and that we talked about last time in Kenya.

    @fortysouth – I think it’s only smart to build a web side to applications these days. The amount of extra work needed to do that is relatively small compared to what is needed on the mobile side (that statement is dependent upon what the app is doing as well though).

  5. How can I resist adding my two rupees worth? ;p

    If we accept that there is a significant difference, not just in technology and infrastructure but also in values and mindset, the worldview between the “developed” and “developing world”, then from my work experience in the field, from the user’s point of view, here are some principles to keep in mind when developing (designing) new products, services or applications (including business models) :

    1. Find the Values Gap – for any particular industry, product or service space, identify the gap between existing solutions available and their value proposition as it maps on to the values/mindset for the BoP market. This exercise will not only provide the filters by which to later evaluate conceptual solution spaces but also the “unmet need” or the opportunity space that can be filled. That is, see what is missing between the real needs in Africa and the solutions for that particular need already existing for other markets.

    2. Maximize constraints; minimize resources – simply due to the challenges of addressing the BoP market successfully, this principle remains constant as a condition and criteria for the design of anything. And taking the thought a step further into the needs of the future of the planet etc this principle becomes even more crucial when making any new product for a global market.
    3. Maximize the freemium – if you’re a startup in this space or just wishing to create a lean organization with minimal impact on the cost of the end result for the BoP market, using as many freely available tools simply makes sense. Web 2.0 exists to be used effectively and efficiently. Also maximizing the freemium aspect of your solution helps ensure greater user acceptance and lowers the dropout rate in the BoP market.
    4. Cross pollination first – The first two principles help identify a general solution space and the design constraints, the fourth is the first place to look for answers. What existing solutions are already out there, whether by a well known MNC like a Tata or a grassroots solution developed on some farm in Uganda. Why reinvent the wheel when you can use it as a precursor or a first prototype from which to build your eventual solution?

  6. 1. Localization – Africa has a massive amount of languages. Even if you’re just shooting for English, French, and maybe Portuguese, if you plan on scaling, it’s something that you need to think of at an early stage. I think you guys are seeing that with Ushahidi. I’ve been burned by it in the past, which is why it’s been the one of the biggest things to focus on with Maneno from the start and even still it is going to get more work as it’s a huge hurdle with both web and mobile.
    2. Know your market – Even if it’s free, you need to know who you’re serving and do that well first before spreading out.
    3. Open your mind – Mobile is probably what defines true viral marketing spread as mobile phones are always on these days as opposed to computers which most of us turn off at night. Once something takes off, it really goes crazy and people left and right will start making suggestions. You have to listen to all of them, even if you discard them. Shutting down the idea stream early on will not allow you to become the next Craigslist.
    4. Web 3.0 anyone? – My guess is that due to the release of iPhone and all its copycats in the US, developers Stateside will finally realize what Africans have known all along in that the slab on your desk or the box at your feet are really limiting what the web can do. Creating new, mobile applications will fuel our next iteration of the web. I think by the end of 2009, most people will declare this as the start of a new wave of innovation as long as the global economies don’t fall further.

  7. miquel, I agree with your point 4 on web 3.0 and I’m optimistic on it (having watched its slow emergence over the past three years) to say that its a wave of innovation that will come out of the developing world, and perhaps for that reason, may not be dependent on the state of the global economy as much as we may assume.

  8. Hash
    The term mobile apps here is misleading-i think. Most the work in Africa still needs to be done on the back end–totally invisible to the user. I am talking infrastructure, security, authentication, dabatase integration with other providers/services, bandwidth etc.
    If done properly the user, by mobile or whatever, should only click one button i.e. press # for balance, * to buy etc.
    Also to make this a “seamless” transaction, real people will need to be hired to work behind the scenes instead of rely on technology.

  9. @Niti Bhan – you and I have talked offline about this now, so thanks for your excellent contribution here too!

    @Miquel – I couldn’t agree more. Really, I think we’re on the front-end of something, and we’ll realize over time that Africa (generally) is well situated to dive right into this new mobile world. The technology gap isn’t nearly as wide as in other areas.

    Your point on localization is extremely relevant, and a great addition, something that I overlooked completely. The question is, do you leave this open for the community? Do you give them editing tools, otherwise it’s too difficult for a central company to setup and maintain the apps for 1000’s of local language dialects.

    @Fimbo – Most apps (web or mobile) really should leave the heavy lifting on the backend. In fact, almost all successful apps, regardless of platform, do this. This leaves a simple and elegant user experience for the main users, and the company earns its money by making the difficult seem doable.

    As far as getting the network providers, and infrastructure all lined up… that might be difficult, especially for your small app creators. However, you’re absolutely correct in the grand scheme of things. Those governments in Africa that make this a priority will be clear economic winners in the next decade.

  10. @hash Yes, I think that you absolutely must have some kind of “wikiness” for localization. While it’s not completely open as you need to have someone moderating the flow of it to maintain consistency, resolve dialectal disputes, etc. it’s really the only way. Well, there is the other way in which you direct hire translators, but technology moves too fast and the expense is too high to make this tenable, especially if you’re doing more than say two or three languages.
    I think Google even shows this as they have more money than the Beatles and yet, they still have their community work on localization, which is one of the reasons that I think their translation system is getting really powerful. I speak two other languages beyond English and that Google’s system even picks up on a number of idioms in some languages, which is impressive.
    I’m trying to figure out some way to do this with Maneno as I keep bumping in to people who are willing to translate, but despite the fact it’s not that much work, getting one person to do the whole job is tricky. I’ve had a lot more luck with one person doing a little and then another person a littler more etc, but still it’s a lot of back and forth. To that end, I’m trying to do something along the lines of what David did and creating a Google doc spreadsheet. It’s just the issue of open sourcing it somehow that freaks me out. That’s going to require more thought and probably building some kind of a simple system for members to contribute to along the lines of having a piece of text that’s untranslated and a link there that allows them to translate it, which then goes to a moderator for approval.
    Now, that’s all fine with a website, but how in the heck do you do that with a mobile app? That kind of two-way interaction doesn’t really exist and the lack of a mouse makes these kinds of things trickier. Being that I’m a web developer first and mobile developer an evolving second, my thoughts always drift back to using some form of web system to manage this, but somehow that is going to have to be overcome which I think is going to be an interesting topic for those much smarter than me in the mobile world to take a look at.

  11. a couple of thoughts on mobile apps.
    1. mobile phones as a platform has only been successful in one area – messaging sending and receiving messages. Any application that goes beyond that in the traditional mobile space has not been very successful
    2. Smartphones are changing the game but i’m not sure if thati in itself is good thing due to propriety designs and creation of “internet walled gardens”
    3.Judging by the moves of industry players, googel w android,nokia w symbian and MS the future of the mobile apps is the web applications
    maybe UI’s will change to accomodate new smartphones.
    4. Africa needs to think out of the box – how about embarcing utility computing perhaps looking at companies like Google and Microsoft as utility companies . govts could for example let companies bid to be the computing utility in the same way cellphone companies bid for bandwidth. In return this computing utilities would provide certain basic services and information on weather and or provide free api access to certain services that would be part of there bid. in return the utilities would be allowed to charge/rent computing services (amazon aws comes to mind).
    while this model may not make sense in developed countries where capacity is great and the technical know how is great for countries with limited skills this would provide an opportunity to ramp up their technology infrastrucure.

  12. in africa laot of information that we take for granted is not readily available. for exmaple simple things such as court records, title and land records, business records, government statistcis are hard to come by and when they exist are generally innacessibel. The hypothetical utility company as part of there responsibility will be charged with digitizing this information and amking it available to the public provide apis (perhaps rent) to allows other developers to tap into this information for there applications.

Comments are closed.

© 2022 WhiteAfrican

Theme by Anders NorenUp ↑