In a recent study of business uses of M-PESA in Kenya funded by the Financial Sector Deepening Trust of Kenya, my former colleague Amolo Ng’weno and I found remarkably low use of M-PESA among formal Kenyan businesses. We collected a catalogue of reasons why, some having to do with entrenched business practices and some relating to limitations of the M-PESA service itself.

So what would a business-friendly mobile money service look like? I’d suggest the following key elements based on the pain points we kept hearing about from business people in Kenya.

Pain point #1: The money went to the wrong account. A customer might mis-type the destination phone number, or biller code when paying the electricity bill. The customer is unhappy: she thinks she paid but her electricity got cut-off, and she is placing irate calls to your staff. Or a customer might pay for goods at your store correctly, and subsequently claim to have made the payment in error. At the very least, your money will be frozen for a while pending some more or less diligent procedure by the mobile payment provider’s call center people; at worst, the payment will be reverted, and you will have been defrauded.

Solution: Checksum digits. Business users should have the option of getting unique biller codes or account numbers which are different from phone numbers, and these should incorporate checksum digits that allow immediate spotting of data entry errors. (Checksum digits must match mathematically the rest of the digits in the account number or biller code, and a mis-typed digit causes the checksum to not match and hence the transaction is immediately identifiable as not valid.) This solution should not be forced on individual users, for whom using their mobile number as the account number vastly simplifies operation of the service. But for business users, why should the corporate payment number be their phone number?

Pain point #2: What’s this payment for? This is a bit the opposite case: a payment went through to a business, but the business can’t figure out who it came from (the phone number does not appear to belong to any of the customers on its database) or what it is for (which of the open invoices from a given customer this payment is intended for).

Solution: Optional reference field. All money transfers should provide an optional field for references, so that senders can put in a customer number (when the money is not sent from their registered phone), an invoice number, or a description for the payment. This is generally done for bill payments, but why wouldn’t this be possible for any business (and some personal) transactions, too?

Solution for extra credit: Enable syntax checks on the reference field. Business customers might be given the possibility of specifying the required syntax on the payment reference field, such that if a paying customer does not enter enough information or it is not of the right nature, the transaction is immediately rejected (“invalid transaction reference, please enter it as [sample syntax]”).

Pain point #3: Customers insist on having a receipt. Customers like having a receipt in their hands, otherwise they fear being powerless in the event of a dispute, or they might need it to show it to their boss, their spouse or whoever. Without access to a receipt, many will keep paying at the company’s window.

Solution: Web-based receipts. Let customers have their receipts, but not necessarily at the point of transacting. The SMS receipt for the transaction contains a unique transaction ID. That, plus the customer’s phone number, might be credentials that can be typed into a dedicated receipts website which then displays the receipt in question. Let people find a cybercafe and print their receipts, if they so want them.

Pain point #4: I want to see all my transactions together. Regardless of how payments are received or made, enterprises want to have a single view of their business. They will not be happy to look up transactions that came or were made via mobile money in one system, and the rest on another system, with a different set of tools and configurations.

Solution: Application programming interfaces (APIs). Mobile money providers must make it easy and safe for businesses to export information on payments received via mobile money in formats that can be used with corporate accounting and enterprise resource management (ERP) systems. I know, easier said than done, but without good integration with existing corporate systems, the business will keep throwing antibodies at mobile money.

Pain point #5: My money is stranded in the mobile money account. Formal businesses tend to like their money sitting in a preferred bank account, and they are not so happy having to hold liquidity balances in a zero-interest mobile money account. Transferring money between the two may take days if they involve interbank settlements. That’s a treasury tax on the business’s operations.

Solution: Immediate liquidity facility. Mobile money providers might work with aggregators who have accounts in all major banks. They can then offer business users quasi-real time conversion of mobile money into bank money, because they can handle all such operations as intra-bank transactions, much like superagents do to offer immediate liquidity to cash merchants.

These solutions would make mobile money more easily acceptable by business users. But they have one added advantage: they can be used to have customers self-select themselves into higher-value business packages, which can drive much more nuanced differential pricing. You want a business customer number with a checksum digit, a syntax check on the reference field, a customer receipt printing capability, a set of corporate APIs and immediately liquidity? Let me tell you about our platinum service…

This is the second post of a two-part series looking at an Oxford University study on the use of mobile money to pay for water services in Africa.

Mobile money payments have the potential to make paying for water a much less burdensome task by nearly eliminating travel to and waiting at congested pay points. Such consumer convenience benefits could improve bill collection efficiency and support wider sustainability in delivering water services. A recent study across Kenya, Tanzania, Uganda and Zambia indicates significant potential but highly uneven performance to date, which was introduced in the first blog and is considered here in relation to mobile network operator (MNO) benefits.

MNOs play an important role in the development and uptake of mobile water payment deployments. In many cases they are the instigators of the option and approach the water service providers. MNOs often hold the upper hand in negotiating the size of customer transaction tariffs, which can be an important factor of adoption. Customers will only switch to mobile payments if the value of using them is greater than the costs of making cash payments. Thus, assessing these companies’ motives and benefits is important to understanding the future of mobile water payments.

MNOs benefit from mobile water payments in two possible ways, as is the case for most applications of mobile money: either through direct profit from fees, or through increased subscribership and reduced customer churn.

Yet in comparison to other mobile money uses, mobile water payments could potentially bring more lucrative MNO payoffs through these. The vital nature of water means payments for it are recurring and regular, offering MNOs a way to become customers’ default SIM card for guaranteed transactions. In this sense they are a good answer to mobile money’s “volume problem”.[1]

Furthermore, intensifying efforts to increase water access for meeting the 2015 Millennium Development Goal, paired with growing opportunities and expectations of “ICT4D” solutions, is driving mobile innovations for improving and expanding water services. For example, in Dar es Salaam, GPRS-enabled water pay points are an alternative payment innovation. Rural mobile payment options are also emerging, such as the Grundfos LIFELINK system in several Kenyan communities that uses M-PESA payments. Beyond payment innovations there are possible markets within other mobile tools: water point surveying applications, which work on networks where available (e.g. Water for People’s FLOW), SMS notification services for water availability (e.g. NextDrop), and attempts at crowd-sourcing via SMS reporting of infrastructure problems (e.g. Maji Matone and the Human Sensor Web). These suggest that basic mobile water payments may just be one of several emerging innovations for MNOs to capitalise on water sector opportunities, though for most, it’s too early to tell how profitable they may be.

The following highlights key findings on the current reality and extent of MNO benefits (direct and indirect) from mobile water payments.

Looking firstly at direct revenue benefits: The MNO fee on mobile water payments is either paid by customers, water utilities, or shared between them. The amount also varies depending on the arrangement: it can be fixed (flat fee for any bill amount), stepped (applied to tiered bands of bill amounts), or a percentage of the bill amount.

This means MNOs take in different amounts in each scenario. For an average-sized water bill paid by mobile money, Safaricom in Kenya would receive USD 0.30, Vodacom would earn USD 0.64 for bills paid in Dar es Salaam, and MTN in Uganda would earn USD 0.09.

Yet net MNO profits are complicated by cash-in/cash-out functions associated with deposits and agent fees. In Tanzania, where a third-party developed the service by integrating MNO and utility systems, the MNO loses some of the revenue to paying the third-party. Figure 1 illustrates this, and it shows how a customer making a cash deposit for mobile money to be used solely for water payment may leave some MNOs without profit, or potentially losing money. Indeed for MTN in Uganda, where USD 300,000 worth of water bills are paid by mobile each month, the revenue generated through facilitation of water bill payments amounts to less than one percent of mobile money profits, which itself brings in less than five percent of total revenue. Similarly, mobile water payments in Dar es Salaam bring in less than one percent of monthly revenue for Airtel and Vodacom. Nonetheless, direct revenue was still identified as a key motivation by some MNOs including MTN in Uganda and Airtel in Zambia.

Figure 1: Estimated MNO profit margin for average mobile water bill payment (USD)

In contrast, some MNOs recognise potentially slim profit margins, and instead view mobile water payments primarily as an opportunity to win new customers and keep current ones (and ultimately have them use higher margin services). Customer loyalty can have a high value for MNOs  – for example, 33% of MTN’s mobile money revenue in Uganda was attributed to a reduction in customer churn.[2] Water payments may be particularly valuable in this capacity as they are essential and recurring transactions. Indeed, Vodacom and Airtel considered this benefit of mobile water payments in Tanzania to be of greater value than direct revenue.

Whether MNOs pursue profits from direct revenue or reduced churn depends in large part on the degree of competition, and it plays out in the customer fee. Across the cases studied, customers are paying widely different fees on mobile water payments, especially when considered as a proportion of average bill sizes (see Figure 2). In Kenya, where Safaricom dominates the market, fees are the highest, amounting to 12.3% of typical bills for Nairobi customers. This relatively high fee may outweigh the value of convenience from mobile payments, especially if making cash payments is not very costly (e.g. in urban areas, a high density of pay-points mean minimal travel time).

Figure 2: Kenyan customers paying water bills via mobile money incur the highest charges in the region.

As expected, some MNOs are dropping these fees to compete more aggressively for customers. Free payment options are offered by Vodacom in Dar es Salaam, and more recently by Airtel in Lusaka and Nairobi. This typically requires water service providers to cover a greater proportion of MNO fees, (as is the case in Tanzania and Zambia), believing it will promote adoption for increased collection and efficiency savings. Water regulators have an influence here too; Zambia’s regulator has prohibited utilities from directly passing on costs of mobile payments to customers.

It is too early to determine the impact of free Airtel bill payments in Lusaka and Nairobi, as both deployments were launched during the course of this study. In Dar es Salaam less than one percent of water utility customers have switched to mobile payments, which Vodacom has offered free since early 2009, with adoption levels constrained by an array of barriers (as identified in the previous post).

These findings and those highlighted in the previous blogpost show overall important potential benefits for all stakeholders- water users, service providers, and MNOs, but the reality of benefits at this point, is unclear.  The commercial motives of MNOs drive a very important piece of the puzzle- as they partly determine customer tariffs and are often the instigator for mobile water payment options. Yet even where MNO competition is leading to lowered customer fees, it is uncertain if this will outweigh other barriers of adopting mobile water payments. What is clear is that there is a need for serious consideration of contexts, operational constraints and behaviours- jointly by MNOs and water service providers (and regulators) for both parties to maximise the benefits from this tool. These wider issues are just as crucial to greater MNO benefits as they are to improving water provision. And with mobile innovations emerging beyond mobile water payments, MNOs may find new markets and opportunities from engaging more with the water sector.

-Written by Ilana Cohen, research team and report authors include Tim Foster, Aaron Krolikowski and Rob Hope. Read the first post of this two-part series here.

[1] Pickens, M. (2011) “Branchless Banking has a volume problem.”CGAP Online Technology Blogpost, Friday, March 11, 2011. http://technology.cgap.org/2011/03/11/branchless-banking-has-a-volume-problem/

[2] GSMA. (2011) Mobile Money for the Unbanked: Annual Report. London: GSMA Association.

The following is a guest post we’re pleased to share by Ilana Cohen, Tim Foster, Aaron Krolikowski and Rob Hope from the Water Science, Policy and Management programme at Oxford University.

Mobile money in Africa offers a promising tool for one of the most enduring and widespread challenges on the continent: the provision of safe water.

The scale of the challenge (and the opportunity) can be seen in the fact that the number of mobile phone subscriptions will exceed the number of people with access to safe drinking water by 2013. At the same time, emerging mobile money options to pay for water services have the potential to increase revenue collection, which is needed for service providers to improve and expand access to water services. On the commercial side, this represents a market in which mobile network operators (MNOs) can potentially build or retain customers to boost transaction volumes, based on regular water consumption patterns.

A research group from Oxford University has evaluated such benefits from mobile water payments in a study of 20 different water service providers across Kenya, Tanzania, Uganda and Zambia. You can read the full report here, and key findings are summarised in two blog posts:

This first post explains the potential value of mobile money to the water sector, and highlights findings of benefits for customers and water service provider (WSPs), with challenges in customer adoption. The second post relates this to commercial interests of MNOs offering mobile water payments.

What mobile money could mean to the water sector’s revenue challenge.

For the water sector in many African countries, mobile money speaks to the key challenge of revenue collection that leaves USD 500 million uncollected a year (equivalent to 0.07% of the continent’s GDP).[i] Poor revenue collection is partly responsible for the fact that 67 million more Africans are without access to safe water in 2008, than in 1990. Infrastructure has simply not kept pace with rapidly growing populations and urbanisation. Unpaid customer bills – averaging up to 40% even among wealthier customers[ii] – cut into operations and maintenance budgets catalysing a vicious cycle of declining services. Such non-payment is unsurprisingly common considering the onerous time and money costs water users usually face when travelling and waiting to make cash payments at limited pay points in underdeveloped regions.

Mobile water payments can potentially bypass these obstacles of cash payments. If the ease of remote payments can enable more timely payments, increased revenue collection should help provide the re-investment needed to improve and expand water services.

Mobile water payments may also be an important tool for those with low or variable incomes, or the unbanked. Mobile money provides a more accessible savings mechanism, and greater flexibility and convenience for making smaller, more frequent payments. For WSPs, mobile payments may reduce administrative costs, or improve accounting of smaller utilities lacking digitised billing systems.

Strong potential benefits and understanding customer adoption

The findings indicate mobile water payments can bring significant customer benefits from time and money savings, and financial benefits to utilities, though customer adoption of mobile payments has been limited in most instances investigated, due to contextual challenges.

The majority of water service providers investigated have one to ten percent of their customers paying via mobile payments (See Figure 1). Uganda’s National Water and Sewerage Corporation leads this pack with 20,000 (10%) of its customers paying USD300,000 per month in mobile payments. The exception to this group is a small-scale and private service provider in Kiamumbi, Kenya (15 km outside Nairobi) with a 76% adoption rate.

Figure 1: Key mobile water payment options across the four countries, and their low adoption rates, with the exception of Kiamumbi.

A household survey was conducted in Kiamumbi to assess customer benefits of mobile payments and understand the motivations, and potential barriers to adopting them. Though Kiamumbi and its piped water network are small (nearly 700 connections and 1,000 households), prior to the mobile payment option, customers could only pay by bank deposit. Yet trips to the nearest bank, 4 km away, can require significant time (66 minutes on average for travel and queuing), and money (USD 0.40 for those who take the bus). In comparison, mobile payments provide an 82% time-savings, requiring only 12 minutes on average (where time is sometimes spent topping up at the agent). And the USD 0.20 fee for the mobile payment is half of the cost for the bus trip[iii].

Unsurprisingly, time-savings was the primary reason respondents chose the mobile option, followed by cost-savings (see Figure 2). In contrast, those not switching to mobile payments cited having to make bank trips anyway or not knowing about the option or how to use it. Some households simply didn’t trust the mobile payments, and the absence of paper receipts was another barrier identified by five percent of respondents.

Figure 2: Time and cost associated with bank trip are key drivers of mobile water payment adoption in Kiamumbi.

The other WSPs investigated hypothesized similar reasons why few of their customers had made the switch to mobile. Currently, limited advertising campaigns have failed to increase customer awareness of the option in several instances. Across all scenarios, demand for paper proof of payment has been a likely barrier to adoption.

Two other key adoption barriers are delayed account reconciliation, and the customer fee for mobile payments. In Kenya, some mobile payments options are not integrated in real-time with utility billing systems, causing accidental water disconnections when payments are made at the deadline. Though utilities have mostly addressed this by delaying the generation of disconnection lists, this initial problem may have impacted customer trust in the mobile option.

The fact that Kiamumbi’s adoption rate was still far higher suggests the barriers in these larger service contexts are greater and more complex. (See the full report for other reasons Kiamumbi’s adoption may be higher.) For example, in large urban areas, frequent pay points may make the time-savings of mobile payments less valuable. This could be the case in Dar es Salaam, where a third-party company offers GPRS-enabled points-of-sale at many city-wide locations.

The cost of the mobile transaction may be another barrier to adoption when it exceeds the costs of traditional payment methods or is very high relative to the total billed amount. This could be a particular disincentive for low-income households who might otherwise use mobile payments to pay in more frequent smaller instalments. The variation in this fee across the study, and how it relates to MNO benefits, is considered further in the following post.

Potential benefits to water service providers and the outlook for mobile water payments

Given low water payment adoption rates, there have been limited benefits for WSP collection rates. However, some have identified cost savings related to bill handling and timelier collection of bills: Uganda’s national provider anticipated savings of USD 420,000 per year from reduced insurance and handling costs, after closing payment offices (though payments can still be made at banks) and enabling mobile payments.

WSPs anticipate further savings if mobile payments can reduce accounting errors, and some say they are seeing fewer disconnections attributed to non-payment. In Kiamumbi, billing records show mobile water payments are 10% more likely to be paid on time compared to bank payments. Though proving a definite causal connection is difficult, customers in Kiamumbi do feel mobile payments are improving their ability to pay: half feel less likely to be disconnected since using them, and one in four feel better able to save for water bills.

On the whole, these findings reveal there is strong potential for mobile water payments to benefit customers, providers, and eventually increase water access as providers reinvest greater revenue. However, the evidence shows the current scale of benefits is limited, where operational and behavioural constraints are barriers to adoption. Mobile money can be an important tool in addressing water sector challenges, but ultimately the scale of its impact depends on understanding and addressing these constraints, and the surrounding institutional and regulatory contexts.

Interestingly, MNOs appear to be one driving force behind mobile payments as they compete for more customers in emerging markets, and have in some cases instigated negotiations with water service providers for a mobile payment option. The next post will consider how this dynamic around MNO benefits is playing out in mobile water payments.

If you are attending MWC in Barcelona next week, make sure to join us for a panel on the subject on Monday 27^th February from 14.00-15.00 (Hall 5, Room 2)

Smart Metering for Better Access to Energy and Water in the Developing

In this session we hope to investigate how mobile operators can get involved in this smart metering opportunity. Even though the market is still nascent, mobile operators stand to benefit directly from increased ARPU and customer retention. The increased mobile phone penetration and the decreased cost of cellular components are spurring innovation from vendors to develop solutions for a better access to energy and water services in developing countries. Several projects in Africa and Asia have been launched in recent months using smart meters (based on GSM technology) to allow consumers to ‘pay as you go’ for energy and/or water the same way they would top up for mobile airtime. We hope that such innovations will also catalyse further socio-economic development for these communities.

[i] Banerjee and Morella. (2011) Africa’s Water and Sanitation Infrastructure: Access, Affordability and Alternatives. The international Bank for Reconstruction and Development/The World Bank: Washington DC.

[ii] Banerjee et al. (2008) Access, Affordability and Alternatives: Modern Infrastructure Services in Africa. Africa Infrastructure Country Diagnostic. The international Bank for Reconstruction and Development/The World Bank: Washington DC.

[iii] Note that these are raw costs, where the averages presented in the full report consider the complexities of multiple transport modes.