Home > Innovation and ICT4D, m4d > Why M-Pesa Outperforms Other Developing Country Mobile Money Schemes

Why M-Pesa Outperforms Other Developing Country Mobile Money Schemes

Why has M-Pesa been so successful in Kenya, yet mobile money initiatives in other developing countries much less so?  Recent Centre for Development Informatics research[1] can help provide a systematic response.

M-money services have two core functionalities.  Registered customers can convert between e-cash and real cash (typically at the physical premises of an m-money agent), and can transfer e-cash from their account to that of another account holder via SMS. They might use this to send money to family members or friends, or to pay a provider – anyone from a taxi driver to a local school – for goods and services.

M-Pesa was launched in Kenya in 2007.  It has grown spectacularly: in mid-2012, there were 19.5 million m-money users in Kenya (83% of the adult population), transferring nearly US$8 billion per year (equivalent to 24% of GDP) – M-Pesa is responsible for more than 90% of these transfers.  Transfers are growing at nearly 40% per year.

It’s not that m-money initiatives in other developing countries have failed: there are an estimated 250m users of m-money services in emerging markets.  Just that they have not – yet – succeeded on anything like the scale of M-Pesa, with Kenya accounting for 30% of all emerging market m-money transactions in 2011.  For example, a recent survey in South Africa found only 16% of respondents with a mobile money account.  In Nigeria, only 3% of adults use mobile money.  And Africa is the lead continent: outside the Phillipines, m-money has been very slow to catch on in Asia. In India, for example, Nokia quit the m-money business in 2012 after two years of failing to build a critical mass.

How do we explain the differences?  University of Manchester research, based on six months of primary fieldwork conducted by Chris Foster, analysed the reasons M-Pesa has grown so fast in Kenya; reasons summarised in the model shown below:

Ongoing support from government – liberalisation of the mobile market; investment in infrastructure; light-touch regulation; facilitation of the initial pilot, etc – combined with strong consumer demand across all strata of society (itself partly fed by the instability and disruption following the disputed 2007 elections).  These drove a virtuous circle:

  • Competition between mobile sector firms pushed them to seek profits beyond the traditional middle-of-the-pyramid; answering the demand from the majority market of the country’s poor.
  • The service was delivered via atomised distribution networks that reached right down into poor urban and rural communities; a network of nearly 50,000 agents by 2012.
  • Those embedded intermediaries – essential in scaling any innovation to reach the base-of-the-pyramid – were given the flexibility to adapt business models, retailing patterns and service offerings so they met the specific and heterogeneous needs of their local customers.  Effective knowledge channels allowed these innovations to filter back up to the lead firms, which then scaled those they found most useful; fuelling yet further growth.

Armed with this model, we can analyse the m-money weaknesses in other emerging markets.  For example:

  • Much lower levels of customer demand (put down to both culturo-institutional factors and more effective functioning of and access to existing financial services) combined with a more stringent regulatory regime are behind the slow growth rates in India.
  • A much smaller number of intermediaries (agents) and a lack of innovation (e.g. to address cash float problems) is restricting growth of m-money in Uganda and Tanzania.
  • Tighter regulation and the much small number of intermediaries has held back expansion of mobile money services in South Africa.

We are not the first to try to understand the different performance of M-Pesa vs. other countries (see e.g. Wolfgang Fengler, Amaka Okechukwu who both also note the value of Safaricom’s market domination).  However, we hope that our model provides a clear and transferable framework for comparison, that can be used alongside more in-depth evidence from other countries to help understand their relative success or failure in mobile money.

If you see ways in which you think the model should be modified – based either on experiences in Kenya or elsewhere; then let us know . . .


[1] Foster, C. & Heeks, R (2012) Analysing policy for inclusive innovation: the mobile sector and base-of-the-pyramid markets in Kenya, paper presented at Globelics 2012, Hangzhou, 9-11 Nov [copy available on request: innov4dev@gmail.com]

  1. 24 November 2012 at 4:59 pm

    I think looking at MPESA distribution (intermediaries and agents) alone without considering other factors such as demographic dividend and interaction with other value chains makes this model too simplistic. I am more in favor of getting more data on economic activity not directly related to Mpesa or Mobile Money schemes and comparing them to see what exactly was different in each case.

    In Uganda it evolved first as a merchant payment mechanism from bars before it spread to P2P transfers. I believe there is too much emphasis on P2P models while ignoring other value chain needs. Uganda probably even has more potential than Kenya if models adapt. Mobile Money can also scale in other markets when business cases are built on realities and needs, not necessarily to ape Mpesa.

    Safaricom faced competitive pressures but the financial sector was also kept napping as they failed to meet the needs of a changing demographic. Ghana and Nigeria are typical examples of countries where there are merchant payment needs the financial sector is not meeting and where innovations like mobile money can be very useful to fill the gap. Regulation seems to be an important factor as you have rightly mentioned and Nigerian mobile money will grow far beyond Kenya if the regulator relaxes its hold.

    This is a huge behavioral change experiment in Africa and the jury is still out. I would trust inferences from data more if we can gather as much of it as possible then analyse.

  2. B.
    24 November 2012 at 5:09 pm

    Chris, you’ll have to factor in that alternative electronic payment methods were already in existence in India and South Africa prior to mobile money. In Nigeria, mobile money has barely started and in a much larger and complex environment, too soon to make conclusion. Again, it’s not just a matter of copying and pasting and thinking that same results will be observed everywhere m-money goes

    You’ll also need to check what mobile money is used for in Kenya versus other places eg.Nigeria, Uganda etc.

    • 30 November 2012 at 10:34 am

      Hi BM!

      it’s not just a matter of copying and pasting and thinking that same results will be observed everywhere m-money goes

      Yes, I would agree with this, and go as far as to say that some m-money implementations have been too focussed on replication and less on adaptation for their specific forms of markets.

      That said, I think our main point in this work is a wider one – that the success of mobile money (and more generally large scale ICT implementations for low income groups) has been influenced by a number of policy factors, and here we are offering some potential directions for policy makers to consider in this respect.

  3. 25 November 2012 at 7:22 am

    Mobile Money in many instances is an evolution – in Kenya it was revolutionary. A perfect storm of massive social upheaval, Kenyan entrepreneurial spirit added to the diagram you have above.

  4. peter wrede
    28 November 2012 at 7:17 pm

    isnt’t it also true that government provided a hugely supportive kick-start by paying teachers’ salaries (or something like that) via M-Pesa thereby creating a critical mass in shortest time? Nowhere else has mobile banking had such a start (and the sad part of it: it can’t be replicated by the private sector alone)

  5. Michael Joseph
    29 November 2012 at 5:15 pm

    I think, as part of the team that launched M-Pesa in Kenya, that this study has completely misunderstood the reasons for the success of M-Pesa in Kenya. It has to do with determination, dedication, passion, trust, brand loyalty, and, most of all, willingness to take the risk of rolling out a massive dedicated and disciplined agency network

    • Richard Heeks
      29 November 2012 at 7:03 pm

      “Completely misunderstood” seems unlikely. That would mean policy, innovation, competition, delivery channels and demand played no role in M-Pesa’s success; which of course they all did.

      Michael’s comment is more appropriate in raising questions about the role of key individuals in driving ICT4D projects to scale; and the personalities and capabilities of those individuals. For the latter, what would be add to: determination, dedication, passion, risk-taking?

      By coincidence, we will soon find out as CDI is just launching a research project on “ICT4D champions”.

      • 30 November 2012 at 2:43 am

        Michael is the former CEO of Safaricom who made it happen. He’s wrong in using the term “completely” misunderstood, but I concur with him that you are looking at in the wrong way. Policy, innovation, competition …. all these are fringe factors. The key thing is this – what need was m-money gonna meet, that would greatly resonate with the people? As B. (#2) & @asemota (#1) highlight – m-money needs to meet a transactional need, one that has no alternative, or which is not accessible / unreasonably structured.

  6. Michael Joseph
    29 November 2012 at 5:16 pm

    And it had nothing to do with teachers salaries – not even today!

  7. David R. Sterry
    29 November 2012 at 11:21 pm

    Learning from M-Pesa isn’t just about replication elsewhere. It’s about how to accelerate the velocity of money globally. How can fees be reduced and accessibility increased when moving money between m-money services, diverse nation states and their currencies?

    • 30 November 2012 at 2:34 am

      Funny thing is David, Mpesa charges more than other providers in Kenya. We feel the pain, but since its agents & market share is huge, we rarely consider the rest. However, simplicity of the solution & iteration of innovations have upheld Mpesa!

  8. 30 November 2012 at 2:28 am

    M-Money is inherently a culture issue. In Kenya, we found it to be an easier way to pay for services. Cheque clearance used to take 5 days when Mpesa was launched, the banks were napping as you’ve mentioned and there were a number of “insecure” ways of sending money back to the rural areas. These are the key needs MPesa came to meet – an easier means of payments in SMEs, a means of “banking” money (not hard cash in hand) and a way to send money upcountry.
    It is also worthy to note a certain myth / rumor going around, which was alluded to in a newspaper story. Michael Joseph, the CEO of Safaricom at the time of development & launch, was known to be very innovation minded. Being an engineer in his previous life, he was hands on, especially with the agents. It is told he gave tough guidelines to his existing distributors – they had minimum registration quotas, which if not met, could cost them their distributorship! And he was known to keep his word! The current CEO doesn’t have that kind of inclination – both of innovation & tough-guy – but is inclined to PR improvements. That is a phase needed now!
    In all, we needed an easier, safer way to send money. @Safaricomltd found a way to realize this need at reasonable terms! And having 80% telecom market share at the time, added to their advantage!

  1. 25 November 2012 at 7:25 am
  2. 18 December 2012 at 10:25 am

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 643 other followers

%d bloggers like this: