Yesterday I woke to exciting news: Safaricom had lowered the minimum size of an M-Pesa transaction and had lowered the tariffs on the smallest transaction tiers. I’m sure that M-Pesa users throughout the country will roundly welcome this news.
And it is exciting. As many on Twitter noted and Peter Gakure-Mwangi noted on thinkm-pesa.com, this opens up the potential for micro-payments at the low end. That’s something that M-Pesa has historically not appeared to be interested in – the addition of the 50 KSH to 100 KSH (about $0.60 to $1.20) tier originally included a tariff of 10 KSH ($0.12). That fee was ridiculous – 10-20% of the transaction amount. But these new fees are much more reasonable, ranging from 5% – 30% (and the higher end of that is a result of bringing the minimum transaction size down even further to a minute 10 KSH). Five percent, while not cheap, is at least in the right ballpark. I’ve argued that it needs to be closer to 2-3% – but this is at least a step in the right direction.
The biggest problem that I see with the fee structure more broadly is that it’s horribly complex. The Safaricom website shows 19 distinct pricing tiers representing various prices for various types of transactions – transfers to another M-Pesa user, transfers to a non-M-Pesa user, and cash withdrawals. Let’s focus on the M-Pesa-to-M-Pesa transfer – then it’s a bit simpler, but there are still 7 different pricing tiers:
So why maintain this level of complexity? Why should the cost of moving 4,999 KSH be so vastly different than moving 5,001 KSH? Who could possibly keep 7 pricing tiers straight, much less 19 when the other types of transactions are included? (Also, why did the new tier get rid of the convention wherein the minimum starts with X1 – shouldn’t it be 51 to 100 instead of 50?)
My proposal would be: make it simple. Make it so simple that everyone knows exactly what is being charged for every transaction. Get rid of tiers altogether. Here’s one idea:
- In-network transfer fee: 1%
- Out-network transfer fee: 2%
- Cash-withdrawal fee: 3%
There are a bunch of ways this could be slightly modified, but I think that its elegance is in its simplicity.
Finally, will these new pricing moves lead to wholesale adoption of M-Pesa at the micro-payment levels? No. Three and five Kenyan shillings is still a huge fee to pay when you’re making numerous small-size purchases during the course of a day. I always think of the “matatu”test – could I see someone paying for a matatu ride (generally 20-50 KSH) with this fee in place? I don’t see that happening – cash is too ingrained, it’s cheaper. But this is an important step to get people thinking about using M-Pesa for more of their transactions – and even if it only happens at the margin right now, it was smart of Safaricom to take a step down this path towards broader usage.
 You could add fee caps so that people aren’t paying 700 KSH on a 70,000 KSH TX if needed; you could continue to have a minimum transaction size so that you don’t go below a fee of .1 KSH.
 Matatus are the 14-passenger form of public transportation most commonly used to get around Nairobi and beyond. They are a force of nature, a bit dangerous and incredibly efficient.
It’s been a busy few months as my wife and I have moved from Kenya back to the US. We’ve settled in the Bay Area so if you’re in the neighborhood, drop me a line.
I still plan to write about the exciting innovations in payments, both those happening here in the US and those in places like Kenya. My experience in Kenya should provide a nice lens through which to look at developments in the space and I hope that you join me in the conversation!
This is the first in-depth post in a series on how mobile payments make money smarter. See the Smart Money Overview for an introduction to the series.
Financial transactions generate a lot of information. But the payer benefits very little and retains control over virtually none of that payment info. Credit card companies have a trove of transaction-based data about payers and they use that data in financial modeling as well as selling aggregated forms of it to advertising firms. But the most information a consumer gets from a financial transaction is a small paper receipt (that will soon be lost or thrown away) and potentially a cryptic line item on the month-end credit card bill (if a credit card was used).
Imagine a world, however, where you as a consumer had full access to your payments and transactions data. That data would include not only the information about your transaction (how much you paid, the exact items you purchased, the date, time and location of that purchase) but also information about the merchant (name of the company and of the sales associate that served you) and the products you purchased (warranty information, consumer reports, any warnings or notices). In isolation, these data may not seem like very much but when aggregated they will be incredibly powerful.
Innovations on the payer side might include:
- Budgeting and mental accounting: Because your phone is with you all the time and it’s able to receive and process payment information in real-time, it will be possible to set up budget alerts that help you stick to a fiscal plan. Does your monthly budget allow for $150 in restaurant spending? Your phone could let you know when you’ve hit $100 and only have $50 left; or if you’ve spent more than half in the first week, for example. There could also be innovative attempts to compensate for the impulse buying that so many find difficult to avoid. If you’ve set up a savings goal for your next vacation (to New Zealand), the next time you’re considering buying that shiny new toy, your phone shows you a picture of your long-term goal (a beautiful New Zealand sunset). For more on the psychology of spending, the NY Times has a good article here.
- Loyalty and rewards: Loyalty and rewards programs are great if you’re really organized. But if you’re like me, you always lose the half-filled punch card and are left to rue the free ice cream cone that could have been. If the mobile phone manages your history of purchases for places that you frequent, you will be eating that free ice cream cone as soon as you earn it. Google’s recently announced Google Wallet includes such capability.
- Smart recommendations: Because Amazon knows what you’ve bought in the past from Amazon, they’re able to recommend items that might be of interest. Same with Netflix in terms of recommending movies based on movies that you’ve enjoyed in the past. That use of past information to provide suggestions will open up whole new experiences and purchases.
- Trends and analytics: All of your payments information can also generate valuable insights on various aspects of your life. Examples are numerous but include: 1) gas consumption, mileage, and costs may provide advanced notice of a mechanical problem as well as a more informed car purchase decision in the future; 2) all those purchases at Starbucks may seem small but in aggregate add up to more than enough to justify the purchase of an espresso machine; 3) an established record of staples purchases may help to foresee the need to buy eggs or toilet paper before you actually run out.
These potential innovations are exciting but perhaps nothing compares to the simple shift in access to these data. Forgot how many hamburger buns you bought for the last BBQ? You can just look it up. Can’t remember how much money you’ve donated over the year for tax purposes? You can just look it up. And so on.
The amount of data created by financial transactions is stupendous and consumers don’t get much value from that data today. When that information is digital, stored securely and opened up to innovators, consumers will reap the benefits.
 My current job involves trying to help people save more of their money in the bank – so this is an area of particular interest. One thing that we hear over and over again from customers is that they need more discipline.
Coins from around the world
One of the things that excites me most about the future of mobile payments is the tremendous amount of services and innovation that the new payment platform will unlock, services and innovation that are difficult to fully imagine today. Take one aspect of mobile payments: the ability to exchange data.
For the entire history of commerce, payments have been static transactions that did not contain or transfer any information beyond the value of the transaction. Payments have been incredibly stupid. Whether it was goods exchanged in barter, cowry shells, coins, or more recently paper bills and credit cards, the payee and payment receiver both retained virtually no information about the transaction other than its value.
Mobile payments hold out the potential to radically change that dynamic. Mobile phones have the ability to store and transmit lots of information embedded within payments. This information will flow in both directions (from the payer to the merchant and from the merchant to the payer) and the opportunities as well as challenges presented by the onslaught of data will be numerous.
In the course of this series, I will explore these opportunities and challenges in more detail.
For now, let’s look at what that data may include (but are certainly not limited to):
- Payer-to-Merchant: payee’s personal identification information, payee’s demographic information, geo-location, time of day, and of course the transaction value.
- Merchant-to-Payer: merchant’s identification information, geo-location, time of day, items purchased, price of each item and total transaction value.
These are the basic building blocks of some incredible innovations. They are also reason for concern and vigilance in managing privacy and information ownership. The combination of those two forces is an exciting area to explore further.
Photo Credit: “money” by fdecomite on Flickr.
Flat revenues and declining profits in FY 2011
Safaricom reported their earnings last week which provides an opportunity to update the numbers on M-Pesa as well as the overall health of Safaricom. As has been well covered locally, Safaricom’s revenues grew while profits fell. Intense competitive pressure from Airtel caused a significant reduction in voice pricing that lowered overall profitability; however, the strong growth in M-Pesa, SMS and broadband revenue compensated for the flat voice revenue.
Note: Data revenue includes both SMS & M-Pesa along w/ internet data
Having dived through the numbers a bit, there are some nuggets that I have not seen elsewhere. Part of that is due to the fact that Safaricom reported annual figures that cover up some of the trends exposed when looking at the numbers for both H1 and H2. Quickly, those nuggets are:
- M-Pesa accounted for 13.8% of Safaricom’s revenues in H2 of their FY 2011. That’s a 23% increase over the 11.2% in the previous 6 months.
- Voice revenue in H2 fell to their lowest level in 2 years. This is not surprising given the slashed tariffs but it is the new reality.
- Revenue was virtually flat between H1 and H2 of FY 2011. This is a big slow-down due directly to the voice revenue fall and something Safaricom must be very worried about.
- Profit before tax and net income both showed steep declines between H1 and H2. This should worry investors even more than the overall declining full FY numbers.
In my last post, I made the case for a percentage-based transaction fee structure in mobile money – especially given that those conducting small transactions are hurt the most. Now I would like to examine a few reasons why M-Pesa isn’t leading the way on lower transactions costs:
- No incentive to do so today
- Lucrative earnings for M-Pesa in current structure
First, M-Pesa is really under no pricing pressure currently to lower the transaction fees. It dominates the mobile money landscape in Kenya with well over 13 million registered users – that’s over 30% of the entire Kenyan population.
Second, M-Pesa is currently making Safaricom a lot of money. In its last financial report (November, 2010), Safaricom reported that M-Pesa earned 5.28 billion Kenyan shillings in the first half of its fiscal year 2011. That represents slightly over 11% of Safaricom’s total revenue for the period. And its share of revenue is growing rapidly: in the first half of fiscal year 2010, M-Pesa’s share of revenue was just under 8% so over 1 year M-Pesa’s share of revenue increased by over 40%.
It will be very interesting to see how those numbers have grown since November. Safaricom releases full fiscal year 2011 results this week and I will report on those when available.
 Safaricom last released M-Pesa uptake data in January so the latest figures are for December, 2010; I expect registered users are substantially higher now.
 M-Pesa’s revenue share in H1 FY 2010 was 7.92% and in H1 FY 2011 was 11.21% making growth of 41.5%.
Current mobile money fees are too expensive for the low end.
Mobile money has the potential to dramatically impact and improve the lives of those living at the “bottom of the pyramid” (BOP). Mobile money is considered safer and more convenient than carrying around idle cash and it also builds a financial record that can be incredibly valuable in obtaining credit for investment opportunities or to handle a crisis. However, mobile money’s promise isn’t being fulfilled today. The current tiered mobile money fees in Kenya make it cost-prohibitive for the BOP to conduct the vast, vast majority of their transactions using mobile money.
Let’s take a hypothetical day in the life of a Kenyan member of the BOP. That hypothetical working day may include a small payment for breakfast, 2 or more small payments for round trip of public transport (if they can afford it), purchase of the goods that they will sell during that day, and then 2 or 3 small purchases from different vendors for that evenings meal (e.g., spinach from one, flour from another, etc.). That’s 6-7 payments for a typical low income Kenyan. (Many people are surprised to hear that poor people conduct more transactions than wealthier people, but it’s true.)
Now, it’s safe to assume that those transactions are relatively small – say an average of $1. At those levels, M-Pesa and other mobile money platforms are simply way too expensive: the M-Pesa fee at $1 is $0.12 which is a 12% fee! It gets worse – any transactions between $1.25 and $437 is a flat fee of $0.37 which is fine if you’re making transfers above $20 (as the fee as a percentage is only 1.85%) but since the vast majority of transactions are far below that, the fee is painfully high. I’ve put together a graph (above) showing the 3 main mobile money platforms in Kenya with amounts and fees translated into US dollars (for ease of interpretation) and their associated transaction fee as a percentage.
To compare: in developed countries like the US, using an electronic form of payment has a transaction fee of between 1 and 2 percent; in Europe it can be even much less than that due to regulation.
All of this to say: M-Pesa and the other mobile money platforms are much too expensive currently for the BOP to use on a regular basis. These mobile money platforms serve an incredibly valuable service (long-distance money transfer). Before M-Pesa money transfer was an insecure, slow and expensive proposition (mostly people sent cash via bus drivers or some other informal means). However, for mobile money to make its next step, to replace cash in every-day life here in Kenya, the mobile money platforms are going to need to reduce their fees substantially.
In my mind transaction fees need to be 3% or below to begin to compete with cash in a meaningful way and unlock the full potential of mobile money – what do you think?
 The CIA Factbook estimates 2010 Kenyan GDP per capita of $1600 (in 2010 USD) which works out to $4.38 per person. This is going to be fairly skewed for the BOP who likely earns much less than the average, especially for a country like Kenya that has a relatively large income disparity. Given that it’s a per person number, however, it undercounts a bit given the relatively large size of Kenya’s non-working population (students don’t earn anything but are still included in the per capita denominator). For ease of the hypothetical, I avoid a specific amount but believe that $3 net per day is probably about right.
 Portfolios of the Poor, a highly recommended book, demonstrated that the poor tend to have a much larger cash turnover than richer people.