Should we be Building SMS or Internet Services for Africa?

Interesting mobile phone

Probably one of my favorite discussions of this trip was entered into after the Uganda Linux User Group (LUG) meeting here in Kampala. It was about whether we should be providing internet protocol (IP) services first, rather than SMS. If cost is the single most important factor for any mobile service aimed at ordinary Africans, then what will it take to move the ball from the SMS court to the IP court? This isn’t just for non-profits to consider, but everyday businesses as well.

Phones that can access data networks have always been in short supply here, so the easy answer has always been to use SMS, just because that’s what people have in their pocket and can use right now. While there are great arguments for either decreasing the costs of SMS, or of moving to IP, the practicality of that was remote due to the costs involved. Either you need a big organization, or a government, who can force the mobile operators to lower their rates on SMS (their cash cow), or you need to have the costs of data-enabled phones to decrease enough that the majority of users switch to them.

There is an argument that says that Grameen’s and Google’s recent deal with MTN Uganda didn’t go far enough in pushing for free, or cheaper, messaging for their new services. Whether you agree or disagree on that matter isn’t relevant if you bypass the argument altogether and provide services via data, which is drastically cheaper, using SMS as the backup.

What a lot of people don’t realize is that for the first time, last year, mobile phones shipped to Africa with data service capabilities outnumbered the simple SMS-only phones that are so prevalent on the continent (Gartner 2009). Of course, this doesn’t mean that there will be a majority of IP accessible phone users immediately, but it is on its way.

Equally important to understand, and a point that increases the momentum of the mobile services over IP argument, is the fact that where there is mobile penetration, there is also available data services. This stands true in Uganda, where MTN says there is 92% GPRS coverage on their network. It’s even true in countries still trying to catch up, like Liberia, where though there are only islands of coverage, that coverage generally comes with data.

Reinier Battenberg, who runs the only local hosting in Uganda, brought up a great point. The fact that Google and Grameen weren’t able to significantly alter MTN’s position on the prices of SMS doesn’t matter. What matters is that Google didn’t offer an IP-based solution for their new Google Trader that they launched. That’s simply unbelievable! It’s doubtful if that type of work would take more than a day for an engineer to implement. Instead of effectively providing an end-run on the strategy around SMS, they just played the same game that the operator wants to play and will win. Something that Google really wants to do is drive people to the web, so why not at least provide web-services for those that can use it? It doesn’t make sense… all around it’s both curious and a questionable strategy.

21 thoughts on “Should we be Building SMS or Internet Services for Africa?

  1. I was investigating a little bit the prices of data access in Sao Tome and Prinicpe and I found out not only the prices are astronomical, the quality is very bad.

    regarding mobile data access I cannot even find anything about data plans or data coverage. The price of a SMS is $0.067 (USD)

    Regarding the Home fixed Internet Access, you pay $300 USD for a monthly 128/256 kbps DSL connection (and that’s the best plan you have)
    If you go for 56 kbps dial-up connection, you’ll pay $0,067 USD a minute

  2. I think this should be considered on a country by country basis when rolling out new products and deciding whether to go sms only or data.
    e.g. due to the prevalence of data enabled phones in SA, then sms only products might limit your penetration, but if you are rolling out the same product in say Uganda, then sms should be a must have, with data being secondary.
    But as technology evolves and we acquire new phone models we will have to move the data route.. the question is how soon or fast are we willing to convert.

  3. Shikoh says:

    In South Africa data is cheaper than sms, take MXit for instance, a simple conversion puts an MB of data to about 13 US cents…

  4. I say we should build internet services – SMS i think is somewhat of a blind alley and even though there are ways to use SMS beyond what its used for now – the cellular incumbents have no interest in developing systems or facilitating applications that undermine their revenue stream. While its in the nature of humans to try to develop things that are cheaper,faster and better

  5. It’s a great question.

    I think the answer and bigger challenge is to build services. How the data is transported is less important as the solution you are trying to provide. I think SMS will have a long term value (especially if prices drop) particularly from a portability/ease of use standpoint. In western markets, I think the advent of smart phones and mobile apps have actually led to increased text message use. We’ve been awoken to the value of mobile services and we’ll continue to want more in either form. I think a lot of services in the future, especially in resource poor settings, would want to handle both. Why would they not?

  6. The right answer is “both”. SMS has reach, data has depth, and any holistic solution will have both. To Google’s play in Uganda, no data layer leaves them open to a Craigslist-type takeover. Yet, as Riba Capital points out, going just IP in Uganda would be foolish.

  7. Safaricom’s recent acquisition of Packet Stream Data Networks and One Communications is a clear indication that the Telco’s are jumping head first into the ISP arena with 4G services. That combined with cheaper smart phones from manufacturers like Nokia means Internet Services are definitely the future.

    Lets face it, we can squeeze only so much out of SMS. It offers only 160 characters, no api, no metadata… nothing, and is dependent on the Telco itself for sending and receiving. Designing any useful apps for such a platform is a tad unpleasant because there just isn’t much to work with – and further, anything successful requires getting in bed with the Telco, who end up being the major beneficiary anyway.

    If one is planning to make money – they’ll have to be patient. A critical mass of data service users is on the horizon, and that is surely the way to go. I just doubt the telco’s are about to let anyone in on the SMS party where champagne has been flowing endlessly for quite a while.

  8. +1 for Wayan @ Inveneo: I don’t see how the answer can be anything but “both”.

    Great SMS based services are being built and released every day. We’ve hardly scratched the surface of what’s possible and there’s a real chance for African technologists to lead on this front rather than follow.

    Pricing and network coverage are the two most important blocking factors for the adoption and success of SMS based services, but even more so for IP based services. Dropping SMS in favor of IP at this point would be totally counter-productive as it would exclude such a huge number people that only a handful of Africans would be able to benefit from mobile services.

    Recap: allow access to your services through SMS *and* IP. Call them ‘light’ and ‘full’ versions if you will and offer different pricing structures if that fits your model. Maximize the potential audience and offer choice.

  9. I too think this needs to be a dual-layer system. Almost everyone should be building for both types of usage.

    The truth is that it’s far easier to create IP-type application hooks than SMS, so I wonder why Google didn’t do that in Uganda with Google Trader. However, it’s an easy thing to fix.

  10. well – actually i have something to add. I think first it seems the voice aspect has been totally ignored – i think in terms of phones the voice aspect is totally under utilized – voice can be used for so much more than just conversations for example to access IVR systems and also voiceXML based websites which are not too difficult to develop and which can support many local languages – voice also is available to probably 100% of the population because you dont have to own a phone you can use anyones phone .

    As for SMS i think SMS can be much more than the box that the telcos have put it into but that requires a change in mindset – what i think is that beginning with shortcodes – i think of shortcodes as analogous to domains now the situation with SMS is almost as though internet domains were rationed.

    Now imagine for a second shortcodes were simple and cheap to obtain and there was a simple service to publish information and make it available so that instead of handing out my phone number to everyone i could simply handout my shortcode for people to text and read my updates.

    So for example my blog could have a shortcode which you could text to get my updates and so on and so forthand i could link other people shortcodes in my posts basically have a sms based web.

    ok im straying a bit

  11. What can you really do with 140 characters of text and no real interactivity?

    SMS is much like a nice little “black and white” TV set, it is a start, but no so much!

    Mobile internet + web capable handsets = something like a “color” TV set, a much better start!

    Of course, telecoms providers do no wish to become dumb tunnels for data, which is what a telecoms industry built more around the internet would do to them. They need to control the applications for more money squeezing out of the already dissatisfied customer!

    Guess who’s king? Customer? Nah, shareholder!

  12. Philbert Muzungaire says:

    I hate to nitpick but perhaps you should rephrase your question to something like “Should we be Building SMS or Internet Services for African clients or customers?.” The original question conveys the sense that you will be giving these products/serices away as charity. I hope that is not your intended meaning. Besides as an African, I am a little tired of “lets do for Africans” sentiments or attitudes. My brother in Bostswana uses Skype, mobile phone SMS, Internet services, etc, according to what he can afford. There is or is not a market depending on country or location.

    Your discussion is welcome nevertheless.

    Best regards,

    Philbert, New York

  13. Great Post & Great Coments !

    SMS and Mobile Internet are different tools that provide difference services and we should use each solution in a way that speaks to their particular strengths. Eventually, Africa will have High GPRS penetration. This still does not mean that a website or webform (or google) is the solution for every problem. I think there a great deal of value in SMS, for example, in terms of data collection, simple request/response, and voice integration . Even people with iPhones still use google SMS.

  14. @Meredith F. Chang – I’d actually disagree with you on this. Both are forms for moving bits of information that allows you read/write access to a database, or hooks into an API. Data networks accessed through your mobile phone are not about loading up websites, at least not in the context of this post, it’s about a different “channel” to pipe information.

    @Philbert Muzungaire – If that was the case, I would have said “Africans” not “Africa” two completely different topics. This is an article pointed at the developers of web/mobile solutions for use in Africa (for profit and not-for-profit), there’s nothing here about charity.

  15. Franklin Nnebe says:

    The cost of data is really the issue and if the focus is on SMS then the operators have really no incentive to upgrade and expand their networks or bring the price of their handsets down. So African service providers should be pushing for IP because that as a side effect will bring down the price of SMS.

  16. I say we should build internet services – SMS i think is somewhat of a blind alley and even though there are ways to use SMS beyond what its used for now – the cellular incumbents have no interest in developing systems or facilitating applications that undermine their revenue stream. While its in the nature of humans to try to develop things that are cheaper,faster and better

  17. John Oxley says:

    The downside of data is it’s lack of coverage. The downside of SMS isn’t cost, or even the lack of structure, but the fact there is no session. An SMS is fire and forget and let’s hope that something comes back. The answer to both these is USSD:

    * Cheap.
    * Coverage everywhere that you can make a phone call.
    * Request/Response is in a session, not completely unrelated packages.
    * Structured menus

    However you don’t get “pretty” applications out of it.

  18. “MTN says there is 92% GPRS coverage on their network”

    Does that mean that all of their subscribers have GPRS enabled on their accounts? And what’s the cost per MB?

  19. I say we should build internet services – SMS i think is somewhat of a blind alley and even though there are ways to use SMS beyond what its used for now – the cellular incumbents have no interest in developing systems or facilitating applications that undermine their revenue stream. While its in the nature of humans to try to develop things that are cheaper,faster and better
    .

Comments are closed.