Mobile Money Annual Report 2011

8/3/2019 Mobile Money Annual Report 2011

1/96

Annual Report 2011


2/96

GSMA Mobile Money for the UnbankedTeam Members

The GSMA

Mobile Money forthe Unbanked Team


3/96

Annual Report 2011

1 Yasmina McCartyManager

2 Paul LeishmanManager

3 Neil DavidsonManager

4 Claire PenicaudCoordinator

5 Seema DesaiDirector

6 Camilo Tellez

Analyst

4 5 6

1 2 3


4/96

Contents

Intoduction

14 Introduction

Chapter 1

718 Is there really any moneyin Mobile Money?

Chapter 2

19 38 Mapping and effectively structuringoperator-bank relationships to offerMobile Money for the unbanked

Chapter 3

3972 Driving customer usage of MobileMoney for the unbanked

Chapter 4

7378 Enabling different paths todevelopment of Mobile Moneyecosystems

Chapter 5

7988 Case study: Mobile Moneyin Paraguay

Glossary

8990

https://www.mmublog.org


5/96


6/96

12 GSMA Mobile Money or the UnbankedIntroduction


7/96

Annual Report 2011

IntroductionOne year on,the Mobile Moneyindustry has doubled in size again

Author: Seema Desai

Over the last 12 months, the Mobile Money industryhas doubled in size; and as the industry has grown,so has our understanding of what it takes to create

a successful deployment. This report containsa selection of important best practices and insightsthat the MMU team have identied.


8/96

34

The Mobile Money industry has blossomed inthe last year, doubling in size for the second yearrunning, and, in May 2011, MMU celebrated thelaunch of the 100th live Mobile Money deploymentin the world. With 88% of live Mobile Moneydeployments in developing markets, MobileMoney is now poised to become a powerful toolfor nancial inclusion; it has the ability to meet theneeds of customers who previously could not accessformal nancial services and had to rely on less safe,less reliable and more costly alternatives.

Africa continues to be the heartland for MobileMoney and home to the industrys most successfulMobile Money service inspiring operators, banks,

governments and other industry players around theworld with the socio-economic and commercialopportunities that Mobile Money services bring.It has nearly 50% of the worlds deployments, andM-PESA, which generated more than 50% ofSafaricoms non-voice revenue and is being usedby over 70% of the adult population in Kenya, hasbecome a key pillar of Safaricoms corporate strategy.1

Increasingly, as the industrys growth becomes aglobal phenomenon, the eye is being drawn awayfrom Kenya and towards other markets. Lastyear, GSMA took its Mobile Money Summit toRio de Janeiro, in order to engage with more Latin

American operators around the opportunities thatexist with Mobile Money. Today, 12% of existingdeployments (and over a quarter of those thatare planned) are in Latin America.2 MMU hascompleted its rst case study of a Latin Americandeployment; focusing on Tigo Paraguay, this studyhighlights some interesting nuances which we hopewill generate further momentum for the growth ofMobile Money within this continent.

This year, the GSMAs Mobile Money Summithappens in Singapore the premier internationalhub in South East Asia, a region that is home to theoldest Mobile Money deployment (SMARTs MobileMoney service, SMART Money, was launched in2003), as well as a number of newer roll-outs incountries such as Cambodia, Thailand, Indonesiaand Vietnam.

As the industry has grown, so has ourunderstanding o what it takes to createa successul deployment

Alongside this rapid growth, MMU has engaged

with a number of Mobile Money providers acrossthe globe in order to synthesise best practice, whichwe have shared through our Working Groups,our website, our publications, and, recently viawebinars. MMU has fully allocated its 5m USD fundfacility in sub-grants to 20 operators across Africa,Asia and Latin America, and these operators havebeen the source of many of the Programmes mostpenetrating insights.

One of the fundamental questions is whethertheres any money in Mobile Money. Do genuinelysustainable deployments exist? The sheer numberof live deployments is a testament to the industrys

belief in the need for - and protability of - MobileMoney services, however, a year ago, empiricalevidence of the sustainability of these services wasscarce. MMU has worked very closely with oneof our grantees, MTN Uganda, to dive deeply intotheir nancial model and assess the key costs andrevenue drivers of the service. The results werepositive and have provided nancial benchmarkdata to the industry for the rst time.

MMU strives to provide the industry with practical,actionable recommendations for how to createsuccessful Mobile Money services. ComplementingMMUs Agent Networks Handbook, which waspublished in last years Annual Report, is a newguide to driving customer adoption of MobileMoney, which is contained in this Annual Report.Additionally, the research that we conducted intohow banks and operators work together can beused world-wide to accelerate the developmentof effective relationships between these partiesto successfully offer Mobile Money services.

GSMA Mobile Money or the UnbankedIntroduction

Mobile Money Developments In Less Developed Countries

20

40

60

80

0

100

Jan 2009 Jul 2009 Jan 2010 Jul 2010 Jan 2011

192221 22 23

25 262627 28

31

45

50 5052

545758 58

64

69

7781

8891

93 9497

97

1 FY 2011 Results Announcement,18th May 2011

2 MMU deployment tracker or this data


9/96

Going orward, MMU remains committed

to helping the industry succeedThe industry has grown quickly, and we have beenbuilding and curating a body of knowledge forthe ecosystem to draw upon. Still, operators facea number of challenges. MMU will be focusing onthe following areas to aid operators deployments:

Regulatory barriers in some countries continue toinhibit the launch and scaling of Mobile Money.MMU will continue to build capacity with mobilenetwork operators, so that they can better engagewith their nancial regulators and help sculptregulatory environments that manage risks such

as money laundering and terrorist nancingwhile enabling nancial inclusion.

Maintaining active and ubiquitous agentnetworks remains a challenge for many MobileMoney deployments. MMU is working hands-onwith operators to identify and solve problemsin their distribution networks.

Bringing a consumer from never having heardof Mobile Money to using the service regularlyhas proven to be a complex marketing challenge.MMU is supporting operators to developmarketing strategies that build awareness ofwhat Mobile Money is, understanding of what

Mobile Money is useful for, and knowledgeof how it works

In certain markets, the industry may need to testnew models in order to exploit more effectivelythe economies of scale. The paper by Ignacio Masfrom the Bill and Melinda Gates Foundation inthis Report highlights why this is important andimagines what might be possible if operatorswere to pursue interconnectivity.

Although some best practices have been identiedin Mobile Money, we have much yet to learn, andthe industry needs to test the reliability of goodpractice across markets. While the MMU teamwill continue to identify and publish learnings,we have also begun to engage more deeply withdeployments around the globe to diagnose keychallenges, develop recommendations and supporttheir successful implementation. We believe thatdoing so will help us to lock-on to crucial challengesmore quickly and devise solutions for them thatwill then drive our agenda over the coming months,ultimately accelerating the deployment of servicesto more unbanked customers around the world.

Over two years, the industry has come a long way.We see more markets approaching the tipping pointof achieving signicant scale such as Tanzania,Uganda and Pakistan and an increasing numberof deployments that are keen to learn from othermarkets and also share what they have learntthemselves. We are more aware of what the barriersare and how to go about breaking those barriersdown. Well continue to build on this strongfoundation, to support the industry and unleashthe full potential of Mobile Money.

Our thanks go to the Bill & Melinda GatesFoundation for their on-going support of MMU,

our Working Group members and, in particular,our grantees, who have allowed us to work closelywith them to develop many of the insights that weshare in this report. Personally, my thanks go to theMMU team for the amount of effort that has gonenot only into the preparation of this Annual report,but also for their hard work and commitment tomaking the MMU Programme, and ultimately theMobile Money industry, realise its enormous potential.MMU Director, GSMA Development Fund

Chapter 1Annual Report 2011

Fiji

Uganda

WestAfrica

CoteDIvoire

Qatar Afghanistan

East Africa

Tanzania

PhilippinesKenya

BangladeshPakistan

Brazil

Thailand

India

Indonesia

Cambodia

Sri Lanka


10/96

56 GSMA Mobile Money or the Unbanked


11/96

Smart Index

Title

Is there really any moneyin Mobile Money?

Mapping and eectively structuringoperator-bank relationships to oer

Mobile Money or the Unbanked

Driving customer usage o MobileMoney or the Unbanked

Enabling dierent paths todevelopment o Mobile Moneyecosystems

Mobile Money in Paraguay

Description

Examines various aspects of theprotability of Mobile Money, basedon the learnings from a deep-dive intothe operational and nancial resultsof MTN Ugandas Mobile Money.

In this piece, MMU shares valuableperspectives based on experiences

from multiple countries on how thecrucial relationship between mobilenetwork operators and banks canwork effectively.

This document highlights the keychallenges that operators have facedwhen it comes to customer activationfor Mobile Money and identiesmarketing tactics that have beeneffective in overcoming them.

Written by Ignacio Mas from theBill and Melinda Gates Foundation.This article identies different pathsto building Mobile Money ecosystemsand develops the commercial case forinterconnection between schemes.

This case study examines the keysuccess factors of Tigos MobileMoney product in Paraguay such asdeep market knowledge, a successfuldistribution network and effective

marketing tactics.

Purpose

To help the broader Mobile Moneyindustry understand the topic ofprotability by taking a closer lookat MTN Ugandas numbers andaddressing key questions relevantto Mobile Money practitioners.

To assist operators and banks that areplanning and/or already working

together to offer mobile nancialservices for the unbanked, providingideas for how roles and relationshipstructures can be rened in orderto promote cooperation.

The help operators drive customerusage, by guiding customers on ajourney from their rst encounterwith Mobile Money to habitual useof the Mobile Money platform.

To help shape industrys thinkingabout the interoperability of MobileMoney services.

To provide lessons for Mobile NetworkOperators in Latin America readyingfor launches and help understandhow this country emerged as a leaderin Mobile Money and what lessons

it offers for the region.

Annual Report 2011

Description o contents o the Annual Report 2011


12/96

78 GSMA Mobile Money or the UnbankedIs there really any money in Mobile Money?


13/96

Chapter 1Is there reallyany money inMobile Money?Author: Paul Leishman

From Afghanistan to Zambia, mobile network operators(MNOs) in developing countries are launching Mobile Moneyservices at a rapid pace. Yet while their enthusiasm to enter thisbusiness is clear to date 100 deployments have been launchedand another 88 are being planned their rationale for doing sois not. Theres no doubt that Safaricoms runaway hit, M-PESA,is protable. But Kenya represents somewhat of an anomaly the perfect coalescence of latent demand, a dominant MNOand a progressive regulator. So the question remains for justabout every MNO outside of Kenya: is there really any moneyin Mobile Money?

MTN UgandasMobile Money is nowcash- ow positive ona month-to-month basis and they crossed thiscritical threshold just14 months ater launch.

Annual Report 2011


14/96


To answer this question, GSMA has studied ourportfolio of MMU fund grantees, which includesrapidly scaling deployments like easypaisa inPakistan, M-PESA in Tanzania and Kenya, and TrueMoney in Thailand; interviewed Mobile Moneypractitioners; and conducted a deep-dive into theoperational and nancial results of MTN UgandasMobile Money, a promising deployment from theEast African country of 32 million where 80% ofthe population lacks access to nancial services.

In an effort to provide a level of depth thats usefulto Mobile Money practitioners, well focus primarilyon MTN Ugandas Mobile Money, but will be sureto put their experience in a global context where

relevant. So before we answer the provocativequestion posed in the title of this article, rst a bitof background on MTN Ugandas Mobile Money.

Launched in partnership with Stanbic Bank inMarch 2009, the service enables customers to sendand receive money domestically and buy airtimeusing their mobile phone; its delivered via anetwork of 1,400 agents; and, most importantly,its growing rapidly, now counting 400,000active customers, processing as many as 385,000P2P transfers per month, and serving as thechannel through which 3% of total airtimeis sold per month.1

Exhibit 1: Growth o active customers and transactionsor MTN Ugandas Mobile Money

Active Customers, dened as any customer that hasperormed a cash-in, P2P transer, cash-out, or airtime top-upwithin 90 days.

P2P transers, dened as a transer o money to a registeredor unregistered customer.

While MTN does have a full roadmap of featuresplanned, weve not made any projections in ourstudy: every insight presented is based on actualdata and has been analysed using our GSMAnancial model.

1 For more inormation, reer to Exhibit 1

0

50000

100000

150000

200000

250000

300000

350000

400000

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

P2P transfers

months

0

50000

100000

150000

00000

50000

00000

50000

00000

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Active customers

months


15/96

Chapter 1

So, is there really any money in Mobile Money?In the case of MTN Ugandas Mobile Money,the answer is yes. The service is now cash-owpositive on a month-to-month basis and theycrossed this critical threshold just 14 months afterlaunch. MTNs peak nancing requirement, orthe amount that they had to nance before MobileMoney became cash-ow positive, was less thanUS$4 million.

Exhibit 2: Financing requirement or MTN UgandasMobile Money

For MTN Uganda, these numbers are exciting.But whats interesting for Mobile Money practitionerseverywhere is exactly how this service becamecash-ow positive. We found that indirect benetsunique to MNOs including savings from airtimedistribution, reduction in churn, and increased

share of wallet for voice and SMS combinedto account for 48% of Mobile Moneys gross prot

to date. We also found that 55% of the costs inthe business to date are variable and step ratherthan xed; in other words, MTNs nancingrequirement has been (and increasingly will beover time) driven by their own customer growth.

Exhibit 3: Gross prot contribution to date (MTNMobile Money Uganda)

Exhibit 4: Breakdown o total, Year-1, and Year-2 costs

(MTN Mobile Money Uganda)

-4000000

-3000000

-2000000

-1000000

0

1000000

2000000

3000000

Month 0 Month 6 Month 12 Month 18

US$

Monthly Revenue, including customer fees for money transferservice, savings from airtime distribution, income from reduced churn,and increased share of wallet from voice and SMS.

Note: the timing of some costs incurred may have been alteredslightly in a way that protects supplier confidentiality but doesnot affect the underlying story.

Monthly Net cash flow, equal to totalmonthly revenue less monthly cash expenses

Cumulative financing requirement, the rolling sum ofmonthly net cash-flow; the lowest point represents the peakfinancing requirement

Monthly Cash Expenses, which includes agentcommissions, customer acquisition costs, technology costs,SG&A costs, and up-front investment in technology (excludesnon-cash items including depreciation of capital assets).

14%

13%

73%

Year 1

66%12%

22%

Year 2

43%

12%

45%

Total

Fixed Costs, including marketing, field agency costs, SIM upgade feesfor non-mobile money customers (assumption for amount attributableto MM), agent handset subsidies, fixed m-wallet provider fees)assumption for up-front investment), agent POS merchandising

Step Costs, including management staff and back-office staff

Variable Costs, agent commissions, SMS fees, SIM replacement,registration commissions, variable m-wallet provider fees, ARPU lossfrom discounting

3%

33%

12%

52%

Money Transfer ContributionAirtime Distribution SavingsRetained ARPU from Churn Reduction

Uplift in Voice/Data Consumption

Annual Report 2011


16/96


How much must an MNO invest inMobile Money beore turning a proft?

The rst chapter of this article introduced MTNUgandas Mobile Money, a service that has turned

an exciting corner into cash-ow positive territory.But the CFO of any mobile network operator(MNO) knows that simply getting out of the redon a month-to-month basis is not enough; hisalternative investment options are usually veryattractive, so he needs to know just how muchis required to scale a Mobile Money service and whether future income will justify the spend.

Unfortunately, theres no generic amount that anMNO in any market, operating with anybusiness model can assume they need to investbefore turning a prot. Until now, SafaricomsM-PESA has provided the industrys only referencepoint; and the best estimates reckon that Safaricomand Vodafone have spent to the tune of US$30million scaling the service so far.2Our teamsrecent analysis of MTN Ugandas Mobile Moneyindicates that theyve spent somewhat less,roughly $10.5 million in total costs and investmentsto date, driving the service into cash-ow positiveterritory on a month-to- month basis. It does meritnote, however, that in its rst 16 months M-PESAgrew twice as fast as Mobile Money in terms ofcustomer registration as a percent of mobilesubscribers (roughly 31% vs. 17% by month 16).

Alas, in the absence of context, top-line investmentgures like these are of limited applicability.For starters, Kenyas population is 36 million so the country is a bad comparable for practitionersin Fiji (population 844,000), India (population1,100,000,000), and most countries in between.Moreover, for better or worse, MNOs in othercountries have not replicated the M-PESA model:in some cases theyve promoted different services,and in others struck different bank partnerships and each of these factors impacts protability.

Finally, and perhaps most important, a successfulMobile Money services nancing requirement

will ultimately be driven by variable and steprather than xed costs; in other words, its difcultto spend like Safaricom unless customers areadopting and using the service.

So instead of asking how much must I invest?,the more relevant question practitioners have

begun asking is what costs will drive my nancingrequirement?. To answer this question, lets againexamine the case of MTN Ugandas Mobile Money.In Exhibit 4, we see that so far, 55% of Mobile Moneysnancing requirement stems from variable and stepcosts, and 45% from xed costs thus, more thanhalf of their nancing requirement has come, inpart, from customer adoption and use.3 We also seethat in their rst year of operation, they incur aninitial urry of xed costs, including investment inthe m-Wallet platform, upgrades to their SIM accessgateway, spending on above-the-line marketing4,and opting to embed their application on all newSIM cards. These xed costs were not insignicant yet as the service grew, they were quickly overtakenby variable costs, including customer registrationcommissions, agent commissions, and per-customertechnology licensing fees. In the second yearof operations, variable and step costs likethese account for fully 66% of the total costsin the business.

Its clear, then, that the nancing requirement fora successful Mobile Money service is driven largelyby variable and step costs but is all the spendingeven worthwhile? That is, can Mobile Moneyservices generate a sufcient net present value

(NPV)? For MTN Ugandas Mobile Money, the signsare promising: if we assume that the service continuesto grow roughly at Ugandas rate of ination andthen include the terminal value in our calculation,the NPV for Mobile Money is positive. Its difcultto say exactly when the cumulative net cash-owcurve in Exhibit 2 will become positive, particularlysince MTN is planning to launch additional servicesthat will surely generate incremental revenue; still,simply based on the foundation theyve laid withtheir domestic money transfer and mobile top-upofferings, its only a matter of time before MTNrecoups its investment.

Its dicult to spend likeSaaricom unless customers

are adopting and using theservice

2 Mas, Ignacio and Radclie, Daniel, ScalingMobile Money (September 22, 2010)

3 And as the service grows, the model willbe predicated even more on variable andstep costs

4 MTN has spent a total o US$850,000 onabove-the-line marketing; this amount isassumed to be slightly skewed to up-rontspending.


17/96

Chapter 1

How signifcant are airtime distributionsavings to proftability?

One of the most important sources of value for mobilenetwork operators (MNOs) who offer Mobile Money

services is the ability to sell airtime using the platform.When a customer buys airtime using Mobile Moneyrather than with scratch cards, operators unlockvalue in two ways. First, they pay lower commissions:the commissions paid to agents for performingcash-in (a necessary step before buying airtime) aretypically lower than the discounts at which MNOssell airtime to the channelalthough the degree ofdifference will vary by market. Second, MNOs saveon the manufacturing and storage of scratch cards.Any savings realised in these ways ow straightto their pre-tax bottom line.

So how big a deal is this? Weve found thatfor successful services, savings from airtimedistribution can be a big deal indeed. For MTNUgandas Mobile Money, this value source hascontributed a total of 12% of their gross protto date. Even though the service is less than a yearand a half old, MTN has still managed to derivesignicant value from their mobile top-up feature:in their best month so far, roughly 3% of totalairtime was sold through Mobile Money at morethan a 9% savings compared to airtime that wouldhave otherwise been purchased via scratch cards.

Beyond Uganda, MNOs are collectively eyeing

or already capitalising on mobile top-up as ameans of reducing their cost of distributing airtime.Safaricom has led the way, apparently selling 19%of its airtime on M-PESA.5 And in the context oftotal protability for their service, this feat has beenimportant: if we assume that Safaricom saves 8%in costs on airtime sold through M-PESA, and assumethat in their last scal year, they sold about $800million in prepaid airtime in total, this suggests thattheyd have generated savings of $12.8 million (notethat these gures are illustrative). By some estimates6,thats more than a quarter of what M-PESA generatedin prots on a standalone basis.

Outside of Africa, mobile top-up has been anequally important value driver for Mobile Money

services and often by strategic necessity. In thePhilippines, where existing domestic moneytransfer alternatives are better than Kenya, bothSMART Money and G-Cash have aggressivelypromoted their mobile top-up services; in Indonesiaand Thailand where regulatory guidelines currentlydont allow customers to withdraw money froman e-wallet, Telkomsel and True Move have bothpromoted mobile top-ups for T-Cash and TrueMoney respectively as an important use ofelectronic funds; and in Fiji, where physicaldistribution of scratch cards to remote areas canbe a challenge, Digicel and Vodafone have bothlaunched with mobile top-up as a core feature.

So how can MNOs evaluate the importanceof mobile top-ups to their protability? The rststep is to identify the size of the discount at whichairtime is sold to the channel: the higher thediscount, the greater the opportunity for MobileMoney to deliver value. Second, an MNO mustestimate the percent of total airtime sales they canreasonably convert from scratch-cards to MobileMoney. And third, an MNO must consider themyriad costs involved in facilitating mobile top-ups.These can include but are not limited to: perpetuitiespaid to top-tier agents on airtime sales for customers

they register for Mobile Money; incentives paiddirectly to frontline agents or customers themselvesto stimulate adoption; and commissions paid toagents for facilitating cash-in (because customerscant buy airtime from an empty e-wallet).

For MTN Ugandas MobileMoney, savings rom airtime

distribution has contributeda total o 12% o their grossprot to date

Annual Report 2011

5 http://siteresources.worldbank.org/AFRICAEXT/Resources/258643-1271798012256/M-PESA_Kenya.pd

6 http://technology.cgap.org/2010/06/07/proo-mobile-money-can-make-money-m-pesa-earns-serious-shillings-or-saaricom/


18/96


How signifcant are churn reductionbenefts to proftability?

If youve ever attended a Mobile Money conference,youve likely heard a speaker tout the potential

benet of reduced churn that Mobile Money canunlock for an MNO. But what you probably haventheard is whether any service has actually deliveredon this promise and if so, whether the subsequentbenets amount to a big or small deal in the overallnancial model.

In our analysis of MTN Ugandas Mobile Money,a service that has turned the corner into cash-owpositive territory on a month-to-month basis, weuncovered a startling nding: in any given month,the churn rate for active Mobile Money customersis negligible. That is, while the churn rate for

regular mobile customers was roughly 4.5% permonth, the churn rate for an active Mobile Moneycustomer was no more than 0.2% over the courseof the three months for which we analysed data.

Exhibit 5: Churn comparison (MTN Mobile MoneyUganda)

This is a dramatic reduction, but the questionremains: does it make much of a difference to theoverall protability of the service? In the case of

Mobile Money, the answer is a resounding yes.Of the total revenue generated to date, churnreduction benets account for 33% and if the servicewasnt delivering this benet, Mobile Money wouldhave barely been out of the red by now. In otherwords, the benet of reduced churn matters a lot.

Alas, there is one catch: not every service wevestudied has generated results as impressive as the

ones described above. Some services report a lessdramatic reduction in churn; some report no changein churn; and some even report a slight temporaryincrease in churn. This variance underscores animportant message for MNOs that launch MobileMoney services on the basis of potential for churnreduction: the benets are real and attainable, butonly for those who execute effectively. That is, theservices that have not realised any churn reductionbenets are those that have registered customerswith no real interest in the service, or been plaguedby bad customer experiences, poorly planned agentnetworks, and half-hearted attempts at creating astrong brand and relevant service offering. Its easyto see, then, why executives in some countries havegone as far as charging internal transfer pricingpremiums to their Mobile Money business units,reasoning that a poorly executed foray intonancial services will do nothing more thanjeopardise existing relationships with valuablemobile customers.

So what does this mean for a Mobile Moneypractitioner? First, it means that execution iseverything. The promise of reduced churn hasbeen realised but only by deployments that arewell funded and have executed effectively.

Second, without considering the benets of reducedchurn, the protability picture is incomplete.Today, many MNOs choose to exclude churnbenets from their P&L or business plan: somedo so because executives are sceptical about whethervariances stem from causation or correlation;others reason that if this service is to be sustainable,it must be on the basis of direct benets alone.The latter rationale is prudent, but when capitalbudgeting season arrives and executives start toask for IRR gures, it behoves practitioners to havethese gures at hand.

And nally, the signicance of churn reductionbenets underscores the importance of trackingthe right metrics. For practitioners to gaugewhether the service is moving the needle on churn,they must rst have a process established, usuallyone in which an external data warehousing teamis engaged, to track the metric. This can be timeconsuming, but given the potential importanceof this metric, its clearly worthwhile.

0%

0.5%

1%

1.5%

2%

2.5%

3%3.5%

4%

4.5%

5%

Month-1 Month-2 Month-3

Active MobileMoney CustomersRegistered MobileMoney CustomersMTN Customers (excluding active and registered MobileMoney customers)

Monthly

Churn


19/96

Chapter 1

How signifcant is ARPU uplitto proftability?

The previous chapter described how signicantchurn reduction benets can be to the protability

case for Mobile Money; and when people talkabout indirect benets, reduction in churn isusually closely followed in the same sentenceby uplift in ARPU (Average Revenue Per User).Having shed light on the important role churnbenets can play in the context of protability,in this chapter well focus on the role of uplift inARPU. But before we answer the question posedin the title of this chapter, lets rst determinewhether uplift in ARPU is even the right metricfor practitioners to measure.

To gauge whether Mobile Money actually causescustomers to spend more, uplift in ARPU wouldneed to be measured over time. But this particulartype of analysis is tricky. First, the average sellingprice for airtime and SMS, and therefore ARPU,in a country varies for any number of reasons ona month-to-month basis, so its impossible to simplyattribute any change solely to Mobile Money.Second, in many cases ARPU gures will alreadyinclude revenue generated from Mobile Money so taking credit again would be inaccurate.

Its clear, then, that uplift in ARPU isnt a perfectmetric. But what, if anything, is? We propose thatthe less catchy, but somewhat more accurate,

phrase of increased share of wallet for voice andSMS is the more relevant metric. By measuringminutes of use and billable SMS events, anMNO can isolate changes in customer behaviour,something thats not possible with an uplift inARPU calculation. Additionally, increased shareof wallet accurately describes just why a mobilecustomer might consume more mobile serviceson their Mobile Money SIM; that is, its easierto imagine a customer who carries two SIM cards,each month spending $3 on one, and $2 on theother, shifting some of her spending to the stickierof her two SIMs. So if we accept increased shareof wallet for voice and SMS as a good metric,

the question still remains: is it a signicant driverof protability?

Unfortunately, our ndings in this department areinconclusive. From a survey conducted in 2009by McKinsey & Co., CGAP and GSMA, we knowthat in the Philippines 44% of Mobile Money userscarry more than one SIM, and 68% report usingtheir Mobile Money SIM as their primary SIM;this is encouraging, but not conclusive evidence thatthis benet is real. In the case of MTN UgandasMobile Money, active customers do consumeslightly more voice and SMS than non-MobileMoney customers, but drawing a solid conclusionhere would be incredibly challenging from adata-mining perspective.

While we havent conclusively pinpointed the

impact of increased share of wallet for voice andSMS in a nancial model, its plain to see that thepotential to reap benets is massive and there aresome steps MNOs can take to position themselvesto do so.

Beyond executing well to ensure customersdo indeed have an incentive to keep their MobileMoney SIM in the phone more often than not(a subject I discussed in the previous chapter),promoting Mobile Money as a method of toppingup is also important. In particular, MNOs havefound success by promoting Mobile Moneyas an option for topping up in small increments,

and topping up after hours when scratch cardsmay be unavailable. For instance, WING,a Cambodian Mobile Money service, has enjoyedsuccess with their mobile top-up feature, and foundthat 33% of top-ups on their system occur outsidetypical store hours, and 70% occur at the US$1 pricepoint, a level at which scratch-cards are a particularlyexpensive as a distribution option.

Annual Report 2011


20/96


How signifcant are direct revenuesto proftability?

So far in this article, weve written about the rolethat indirect benets play in enabling a mobile

network operator (MNO) to turn a prot from aMobile Money service but what about the mostobvious source of value, direct revenue fromcustomer fees? After all, this is often the singlesource of value upon which MNOs evaluate thebusiness case for Mobile Money.

For MTN Uganda, who currently offer domesticmoney transfer and mobile top-up services, directrevenues include fees to send money, and fees towithdraw money from an e-wallet. To date, thesedirect revenues, less commissions paid to agents,contribute 52% of total gross prot for the service.Its clear then, that this is an area of the businesscase not to be neglected. So how can MNOs ensuretheyre well positioned to fully capture this sourceof value? Well in the case of MTN Ugandas MobileMoney, one decision has had more of an impactthan any other: enabling P2P transfersto unregistered recipients.

Uganda is a fragmented mobile market: accordingto Wireless Intelligence at time of writing, MTNholds 44%, Zain, Warid and Uganda Telecom eachhold roughly 18%, and Orange holds 3% marketshare. So its not surprising, then, that when MTNlaunched the service, they made sure customers

could send funds to recipients on any network.To date, 38% of P2P transfers made using MobileMoney have been from a registered customer toan unregistered recipient; and this use case hasgenerated 45% of total revenue (and even more ingross prot). Two things are striking about this data:rst, the overall number of P2P transfers tounregistered users is quite high, which suggeststhat had MTN not offered this option, they likelywould have left some revenue on the table. Second,P2P transfers to unregistered users are morelucrative for MTN than P2P transfers to registeredusers (i.e. 38% of transactions are generating 45%of revenue). This occurs because MTN charges

customers a premium 7% for low and 94% forhighest value transfers to make a transfer to anunregistered recipient, and the commission paidto agents remains the same. Thus, by enabling P2Ptransfers to unregistered recipients, MTN not onlyexpands the base of potential users for their service,they also generate a signicant amount of revenue.

But not every MNO allows P2P transfers tounregistered recipients: some reason that by doingso, they are forfeiting potential net new mobilerevenue from recipients who, if they want to receivemoney, have no choice but to activate a SIM fromthe MNO in question (and then, as the theory goes,start to use this new SIM for mobile services, too).But this walled garden logic is risky: Mobile Moneyis a service that is predicated on network effects,and particularly in countries with fragmentedmobile market share, the closed model presentsan insurmountable customer experience barrier

to adoption, ultimately making it difcult to scalethe Mobile Money service. And if a Mobile Moneyservice cannot scale, its sustainability becomesquestionable so in the end, any benets of netnew revenue will be short lived.

Its clear, then, that direct revenues are a signicantvalue source, and mobile network operators havean opportunity to maximise them by enabling P2Ptransfers to unregistered recipients a feature that,coincidentally, is just what customers in Kenya,Uganda and other successful Mobile Moneycountries have demonstrated that they want.

Exhibit 6: % o

transactions tounregistered customers

Exhibit 7: Money transer

revenue

55%

45%38%

62%

Average volume of transfers to registered customersAverage volume of transfers to unregistered recipients

Revenue from transfers to unregistered recipientsRevenue from transfers to registered recipients


21/96

Chapter 1

How can an MNO manage costs toachieve proftability?

When most people hear the phrase to turn a prot,we need to manage our costs, they usually take

it to mean to turn a prot, we need to reduceour costs. But when it comes to Mobile Money,practitioners have found that some costs can bedone away with more easily than others. So thetrick, then, is to understand which are strategic(and must be protected), and which are discretionary(and can be curtailed).

But before we begin our evaluation process, letsrst briey take stock of the costs (and there aremany) that are involved in launching a MobileMoney service. Before launch, MNOs incura series of technology costs, including investingin an m-wallet platform, upgrading their SIM orUSSD access gateway (in most cases), and decidingwhether to embed their application on all newSIMs and in most cases consequently upgradeto a larger card (while this isnt a cash outlay at rst,its a decision of major nancial signicance).The next tranche of costs are go-to- market related,and include recruiting and paying for managementand back-ofce staff, training and merchandisinga network of agents, and designing and launchingabove and below-the-line marketing 30% campaigns.Most of the costs identied thus far carry 7% onafter the service has been launched, but the day aservice goes live, a third set of costs come into play:

ongoing costs. These typically include cash-in/cash- out commissions for agents, SIM cards,starter packs and agent registration commissionsfor customer acquisition, and internal transfer feesfor using SMS services or selling airtime at adiscount. For a full breakdown, refer to Exhibit 8.

So which of these are strategic and which, if any,are discretionary? Unfortunately, answering thisquestion is not as simple as sorting costs accordingto size. If we look at the drivers for MTN UgandasMobile Money, we nd that highly strategicoperational activities things like building andmanaging an agent network, or providing great

customer care are comparatively inexpensive.Since launch, 7% of Mobile Moneys total costshave been on building and managing their agentnetwork6, and 4% has been on back-ofce customercare.7 And while its true that Safaricom spendssomewhat more on these particular activities, andhas beneted from an agent network of industryleading quality, the insight is still applicable:these activities are routinely touted as strategicimperatives for any successful Mobile Money

service but for MTN, theyve cost a pittancecompared to the amount spent on technology8 (30%)or customer registration commissions9 (12%) to date.

Exhibit 8: Detailed breakdown o costs

So if these activities deliver such good value formoney, why do some practitioners have a difculttime getting budget to do them properly? In manycases, this stems from the fact that highly strategic,nancially insignicant costs often require acommitment to spend in advance of having anyindication of whether the Mobile Money servicewill be a success. For instance, MTN had to committo a xed monthly contract with a eld marketing

agency ($623,000); pay for and train their dedicatedcall centre representatives ($440,000); and designand fund an above-the-line marketing campaign($850,000) all prior to launching their service. Eachof these activities has been instrumental in MTNUgandas success, and their decision to investaggressively in them ultimately stemmed fromtheir condence that the service wouldbecome a hit.

6 Includes handset subsidies, agent POSmerchandising, and feld marketing agencycosts

7 Includes total cost o back-ofce sta

8 Includes cost o m-wallet platorm andmonthly charges, SIM access gatewayupgrade and monthly maintenance charge,and SMS communication ees

9 Includes commissions paid rom MTN toagents ($1.33 per registration).

7%

28%

12%

Agent network costsHandset subsidies POS merchandising field marketing agency contract

Technology costs

Up-front investment in m-wallet solution SIM access gateway upgrade recurring feesfor m-wallet solution recurring fees for SIM access gateway upgrade maintenanceSMS communication fees (internal transfer price)

ARPU loss from discountingTotal airtime bought through mobile money at a discount multiplied by discount rate

Selling expenseMarketing and advertising

G&AManagement staff back office staff

Cash-in/cash-out commissionsMoney transfer commissions agent airtime commissions

Customer acquisition & registration costs

Customer registration collateral registration commissions SIM swap cost incrementalSIM cost for upgrading to larger card

30%

3%

8%

12%

Annual Report 2011


22/96


But thats not to say all of their spending has beenstrategic; some costs were discretionary, and potentially

could have been substituted for less expensive,equally effective alternatives. For instance, MTNrecently introduced an airtime bonus for customerswho top-up using Mobile Money an incentive manyMNOs have used in an effort to encourage customersto top-up using their e-wallet. But this tactic wasparticularly costly since it negates a big portionof the savings realised from eliminating discountspaid to dealers.

Moreover, like they have in other markets, MTN haspursued a strategy of aggressively registering newcustomers in Uganda. In practice, this has meantregistering more inactive customers (552,213) thanactive ones (421,254). And this strategy has beenexpensive: MTN has spent a total of $1.3 millionon registration commissions and new SIM cardsfor customers that have not performed a singlerevenue-generating transaction.

How can MNOs ensure their tari andcommission models are well designed?

For a Mobile Money service to scale and achieve

protability, its critical to have well designedcustomer tariff and agent commission models.So how can MNOs ensure their tariff andcommission models are well designed? Hereagain, MTN Ugandas Mobile Money exempliessome key insights.

If the Mobile Money customer tariff model looksfamiliar to you, thats probably because youveseen it in action before: in structure, its a replicaof Safaricoms M-PESA. And as Ignacio Mas notedin the 2009 Mobile Money for the Unbanked Annualreport, this tariff structure (and the way its takento market) works for a few reasons: its simple and

transparent, customers are not bound by minimumbalance requirements or prohibitive deposit fees,and it offers customers an ability to send moneyto non-customers.10 Its inevitable that MNOs willinnovate and trial new models, but the designfeatures listed above can be considered prerequisitesfor an effective tariff model in any environment.

MTN Mobile Money rates

It also merits note that MTN Ugandas customertariff model grants customers minimal leewayto defraud the operator of prospective directrevenues. That is, given that the P2P transfer feetypically accounts for less than half of the totalend-to-end cost of sending money using the service,customers have little incentive to perform a directdeposit. Moreover, MTN has structured its tarifftiers in such a way that there is no opportunityfor a customer to reduce their fees by splitting acash-in or cash-out into multiple smaller tranches.

But its not just MTNs customer tariff model thatmerits attention. Their agent commission model

has been thoughtfully designed, too. The articleNeil Davidson and I wrote for the 2010 MobileMoney for the Unbanked Annual Report detailsmost of our thinking on agent incentives, but itsworth briey noting here how MTN espouses somekey principles.

First, MTN pays Mobile Money agents a commissionfor every activity that they perform, even thoughMTN may not charge customers a fee directly foreach one. For instance, even though MTN doesnt

MTNs decision to investaggressively in marketing,

agent monitoring, andcall centre sta ultimatelystemmed rom theircondence that the servicewould become a hit

Min Max

Activity

Loading Money

Sending Money

To registered user

To non-registered user

MTN or Local network

Withdrawing

By registered user

By non-registered user

MTN or Local network

Buying airtime

Daily transaction limit

Transaction Tiers

5,000

5,000

5,000

30,0001

60,0001

125,001

250,0001

500,001

5,000

30,001

60,001

125,001

250,001

500,001

5,000

5,000

UGX 1,000,000

1,000,000

1,000,000

30,000

60,000

125,000

250,000

500,00

1,000,000

30,000

60,000

125,000

250,000

500,000

1,000,000

1,000,000

1,000,000

Charge

0

800

1,600

2,000

3,700

7,200

10,000

19,000

700

1,000

1,600

3,000

5,000

9,000

0

0

10 Ignacio Mas: Good Service Design Featureso M-PESAs Money Transer Service. 2009MMU Annual Report.


23/96

Chapter 1

11 This decision also helped drive customergrowth.

charge customers a fee to cash-in, they do provideagents a commission for providing this service inrecognition of the time and cost involved. Of course,while MTN take a temporarily hit by subsidisingcash-in, the fees collected from an end-to-endmoney transfer (which includes a cash-in, a transfer,and a cash-out) do exceed the correspondingcommissions paid. All told, the margin MTN earnsfor a typical end-to-end P2P transfer (excludingvariable technology fees) to a registered customeris just north of 50%.

Second, while MTN may pay agents for both cash-inand cash-out, they deliberately pay a highercommission to agents for facilitating cash-outthan they do for cash-in

. This stems from thesimple fact that cash-out agents have a highercost of restocking their inventory of physical cashthan cash-in agents do for restocking their inventoryof e-money. As such, cash-out agents must becompensated accordingly.

Third, MTN recognised that to keep agents engagedin the period following launch when transactionvolumes are typically low, it would be importantto provide them with a different source of revenue.To this end, they have provided agents with acommission for every customer that they register.11

Thus, in the early days following launch, Mobile Money

agents earned money by registering customers;as the service scaled they increasingly earned theirmoney from facilitating cash-in and cash-outtransactions for customers.

Appendix A: Resources

GSMA Financial Model

The GSMA Financial Model is an excel tool thatpractitioners can use to develop a comprehensiveview of the protability of their Mobile Moneyservice. The model generates a P&L statementthat is based on a series of user inputs, includinginvestment, direct benets, indirect benetsand costs.

GSMA Metrics Dashboard

The GSMA Metrics Dashboard is an excel tool thatpresents practitioners with an easily digestiblesummary of their operational metrics that mattermost. Existing and future customers of Comviva,Fundamo, Sybase 365 and Utiba can integrate the

Dashboard as a reporting feature free of charge.

To receive a copy of the GSMA Financial Modelor the GSMA Metrics Dashboard, send an emailto [email protected].

Annual Report 2011


24/96

1920 GSMA Mobile Money or the UnbankedMapping and eectively structuring operator-bank relationshipsto oer Mobile Money or the Unbanked


25/96

Chapter 2Mapping andeectively structuringoperator-bank

relationships tooer Mobile Moneyor the UnbankedAuthor: Neil Davidson

Foreword

We are living in an era of unprecedented change.One of the most transformational of these changeshas been the inuence of the mobile phonewhichhas become one of the most commonly usedtechnologies on our planet.

We continue to see the ways in which people usetheir mobile phones grow and change. One ofthe most important of these has been in nancialservices, an area that will have a signicant, positiveimpact on the global economy. When peopleaccess nancial services applications through theirmobile phones, they become members of the digitaleconomy, opening up a new set of opportunities,

particularly for the unbankedthose individualswho are completely outside of the bankingsystem today.

For those of us in the mobile nancial servicesecosystem, Mobile Money represents both an

opportunity and a responsibility. The businessopportunity is clear, but with that comes a

responsibility to work together as an industryto leverage each others strengths in order toreach those currently excluded from formalnancial services.

One of the critical pieces necessary to make mobilenancial services work is the relationship betweenmobile network operators and banks. To be effective,this needs to be a win-win relationship.

Mapping and effectively structuring operator-bank

relationships to Offer Mobile Money for theunbanked by the GSMA shares valuableperspectives based on experiences frommultiple countries on how this relationshipcan work effectively.

I invite you to read this interesting publicationand hope you can make practical use of itslessons learned.

Tomasz Smilowicz, Global Head o Mobile SolutionsCiti, Global Transactions Services

Annual Report 2011


26/96

2122 GSMA Mobile Money or the UnbankedMapping and eectively structuring operator-bank relationshipsto oer Mobile Money or the Unbanked

Executive summary

In the past several years, both banks and mobilenetwork operators have moved aggressively to

offer mobile nancial services to the unbanked.For banks, Mobile Money for the unbanked is a wayto serve a vast swathe of customers who wouldotherwise be out of the reach of costly branchinfrastructure; for operators, Mobile Money representsan opportunity to differentiate themselves fromtheir rivals. To offer Mobile Money, banks andoperators need to work togetheryet negotiatingagreements to do so can be contentious andtime-consuming: our research suggests that earlyattempts to forge these agreements took banks andoperators a full year, on average, to negotiate.

In this article, we introduce the idea of the businessowner: the bank, operator, or third party thatassumes the bulk of the nancial risk of offeringa Mobile Money service. The business ownercontracts with other entities to undertake the activitiesin the Mobile Money value chain it chooses not tooperate itself. We take a close look at these activitiesand evaluate which partya bank, an operator,or a third-partyhas the most relevant assets andcapabilities for each task. In general, we nd:

Operators have a widely recognised andaccessible mass-market brand, which most bankslack. However, banks are more experienced ineducating their customers and persuading them

to consume a service that, unlike airtime, theydidnt already know they need.

Operators know how to build networks ofindependent retail agents and can leverage thesenetworks to serve as cash-in/cash-out pointsfor a Mobile Money service. Banks, particularlythose with branches in rural areas, are ideallysituated to support agent liquidity.

Both banks and operators have experience runningtransactional platforms, although in practice,the platform itself is usually built by a third party.

Given existing relationships, banks are betterpositioned to engage with regulatory authorities.But we discuss the signicant tensions that canarise when a Mobile Money service with anoperator as its business owner is viewedas bank-led by the regulator.

We also discuss how agreements can be formalisedand value allocated. We point out that while

operators need not work with just one bank, it isharder for banks to work with just one operator.We consider what functions are easily outsourcedto another entity by the business owners andwhich are not. And we discuss the pros and cons ofcomplex agreements that allow two or more partiesto share business ownership of the Mobile Moneyservice. Regardless of their complexity, we highlightthe three hallmarks of successful agreements:clarity about roles and responsibilities, a win- winproposition that extends into the future, and explicitgovernance structures.

Two appendices are included at the end of the

article. The rst is a tool that operators and bankscan use as a framework when looking to structure(or re-structure) their engagements. The second is apair of case studies showcasing engagement modelsbetween banks and operators in Kenyaand Pakistan.


27/96

Chapter 2

Introduction

It is impossible for a mobile network operator tooffer Mobile Money without a bank: at minimum,a bank must hold the deposits which back theelectronic value stored in customers and agentswallets. Conversely, it is impossible for a bank tooffer Mobile Money without an operator: at minimum,an operator must provide the data channel whichallows customers and agents to initiate transactionsusing their handsets.

But between these two extremes there is a verywide variety of ways for banks and operatorsto work together. Telenor Pakistan and TameerMicronance Bank have together created a virtualorganisation to run their easypaisa service, nely

sorting roles and responsibilities and allocatingthem between the partners; more typical is fora bank to handle two or three functions and theoperator to take on the rest. Sometimes, thesearrangements are formalised with contracts andservice level agreements, with one party agreeingto offer a service or services to the other for a fee.More rarely, operators and banks may enter into ajoint venture, or nd some other way of sharing inthe risks, and the rewards, of offering Mobile Money.

This diversity of options, paired with the necessityof striking some kind of deal, can make the processof negotiating contentious. In one African country,

a proposed Mobile Money service has been stalledfor more than a year while an operator and a bankhave debated the nature of their relationship. This isnot atypical; our research indicates that, on average,negotiation between a bank and an operator seekingto work together on Mobile Money takes twelvemonths to complete. Even when negotiations areconcluded, it can leave one or both parties uncertainwhether theyve hit on the operating model thatallows them to build a Mobile Money service mosteffectively, and to capture an appropriate share of thevalue thats created in the process. Such uncertaintiescan reduce the effectiveness of banks and operatorswhen developing and rening a service that truly

meets the needs of the target market.

To shed light on these issues, we seek to answer afew fundamental questions about the relationshipsbetween banks and operators in this article:

What are the respective strengths that mobileoperators and banks bring to Mobile Money?

What are the activities that need to beperformed to offer Mobile Money, and whichparty (a bank, an operator, or a third party)is best equipped to perform each?

What are the different ways that banks andoperators can engage with each other?

How can banks and operators structure,or restructure, their agreements to reducefriction and improve the service that theyoffer to their customers?

We posed these questions in a series of interviews

to dozens of executives at banks and operators inAfrica, Asia, and Latin America. We spoke withrepresentatives from multinationals and fromcompanies with operations in just one market; withstrategists and with line managers; with those onlycontemplating Mobile Money and with industryveterans. We are grateful to them for sharing theirinsights and experiences, which form the backboneof this article.2

Surveying the landscape

As of May 2011, there are 100 live Mobile Moneydeployments in low- and middle-income countries

that target the unbanked.3

In each, it is possible toidentify a business owner, which we dene as theentity which assumes the bulk of the nancial riskof offering the service. The business owner contractswith one or more parties to provide certain services.If successful, the business owner captures the residualprots from the venture after all other parties havebeen paid.4 In this article, we will identify thebusiness owner of various deployments rather thancharacterising them as bank-led or operator-ledsince although these terms are widely used, they arevaguely and inconsistently dened.

In principle, a bank, a mobile operator, or a third

party or some combination thereofcan serveas the business owner. Today, we see the following:

In the large majority of cases, the mobile networkoperator acts as the business owner, contractingwith one or more banks to provide servicessuch as oat holding and regulatory engagementand compliance.

In operator-bankpartnerships, each entity has

to have the trust to let theother do what they do best.

Nadeem Hussain, CEO,Tameer Microfnance Bank1

1 Tameer Microfnance Bank, which is partiallyowned by Telenor Pakistan, oers a MobileMoney service called easypaisa with Telenor.For more about the Telenor-Tameer partnership,see the appendix.

2 We are grateul to Rambert Namy andAlexander Boeller o Sorecom and toAmitabh Saxena or their work researchingthis article, and to Chris Bold or supplyingimagery.

3 See the Mobile Money or the UnbankedDeployment Tracker (http://www.wirelessintelligence.com/mobile-money/unbanked/) or a list.

4 Economists would call the business ownerthe residual claimant: the entity with a claimon profts ater all costs have been paid andall debts have been repaid.

Annual Report 2011


28/96

2324

In a handful of cases, a bank or bank subsidiary actsas the business owner, contracting with one ormore mobile operators to provide services suchas access to short codes and the USSD gateway.

There are two services offered by a partnershipbetween a bank and an operator in which the twoparties share in the risks, and the potential protsfrom, Mobile Money: easypaisa and M-KESHO(a third, MTN Banking, is being absorbed backinto Standard Bank after a number of years asa joint venture co-owned by Standard Bank andMTN in South Africa).

Finally, there are a handful of third parties, likeSplash in Sierra Leone, that act as the businessowner, contracting with both banks and mobile

operators to provide services required.

In this article, we discuss a wide range of MobileMoney services for the unbanked, regardless of theirbusiness owner. And we include services that rangefrom the basic (bill payments) to the sophisticated(savings, insurance, and credit).

Excluded from our analysis, however, are mobilenancial services that are primarily conceivedby banks as channel extensions, giving theircustomers, who are by denition already banked,a new way to interact with the bank, complementingexisting channels such as branches and internetbanking. Thats not to say that such services areunimportant, for banks or mobile operators. In fact,they can be popular with users, a competitivedifferentiator, and protable for banks and operatorsalike. But they are different enough from servicesthat target the unbanked that we have chosen notto discuss them here. (In the next section, we discussthe key feature that distinguishes an unbanked-focused Mobile Money service from a channelextension: a network of independent agentsat which customers can cash in and cash out.)

Why banks and mobile network operators are interested inMobile Money or the unbanked

In a 2009 study commissioned by the GSMA andthe Consultative Group to Assist the Poor (CGAP),McKinsey & Co. estimated that there will be 1.7billion unbanked customers with mobile phonesby 2012 and that up to US$5 billion in directrevenues can be earned by serving this segmentbetween 2009 and 2012.5 Mobile operators andbanks have obvious, but distinct, strategic interestsin serving this market.

For banks, Mobile Money is a way to serve a vastswathe of customers who are otherwise out ofreach. Generally speaking, the low-income segment

cannot be protably served using the traditionalbanking model, in which bricks-and-mortar branchesare the primary point of contact between customersand the nancial institution. Thats because it israrely economical to build and operate bricks-and-mortar branches, with their high xed costs, wherethe poor live: even if such a branch were busy allthe time, the fees the bank would have to chargetheir clients, relative to the size of those clientstransactions and/or deposits, to cover the branchscosts would exceed customers willingness to pay.6

And this problem is exacerbated in rural areas,with low population density. In contrast, MobileMoney services allow users to cash in and cash out

at a network of independent agents, leveragingexisting infrastructure to serve customers morecheaply than in a bricks-and-mortar branch.7Moreover, customers can then move value (whetherit is to pay bills, send money to a relative, or performsome other transaction) by issuing commandsdirectly from their handset, here again leveragingexisting infrastructure to further bring down thecost of serving poor customers. As such, MobileMoney allows banks to prot from helping to servea market they might otherwise have to forsake.

5 See Understanding the UnbankedCustomer and Sizing the Mobile MoneyOpportunity by Paul Leishman (http://www.gsmworld.com/ documents/mmu_2009_annual_report.pd).

6 This makes the achievements o pro-poorbanks like Equity Bank (Kenya), Grameen Bank(Bangladesh), and BRI (Indonesia), whichhave managed to establish branches even inlow-income areas despite these challengingeconomics, all the more impressive. See TheEconomics o Branchless Banking by IgnacioMas (http://mmublog.org/global/article-rom-ignacio-mas-the-economics-o-branchless-banking/).

7 See Scaling Mobile Money by Ignacio Masand Daniel Radclie (http://papers.ssrn.com/sol3/papers.cm?abstract_id=1681245).

GSMA Mobile Money or the UnbankedMapping and eectively structuring operator-bank relationshipsto oer Mobile Money or the Unbanked


29/96

Chapter 2

How Mobile Money or the unbanked ts into a banksbroader mobile strategy

It would be unusual or a banks only use o the mobilechannel to be oering a Mobile Money service or theunbanked. More commonly, banks rst seek to exploit themobile channel as a new way o serving their existingcustomers. By allowing customers to check their balances,view transaction reports, and move money betweenaccounts, banks oer customers a value added serviceand realise operational savings (when customers chooseto interact with the bank by mobile vs. through moreexpensive channels, like telephone or a branch). They mayeven earn additional revenues i customers are willingto pay to use the mobile channel.

Banks who participate in the value chain o a Mobile Moneyservice or the unbanked typically see that initiative asdistinct rom their use o the mobile channel to better servetheir existing customers. Since ew banks target the samecustomers in their core business as in the Mobile Moneyservice or the unbanked, the potential or cannibalisationis typically low.

For operators, Mobile Money does not usuallyrepresent an opportunity to serve a new marketsegment; instead, it allows them to cross-sella new service to customers whom they alreadyserve (i.e., their own subscribers) or compete for(the subscribers of other mobile network operators).

Given the increasing competition in developingcountries among operators for share of the mobilebusiness, and the increased propensity of customersto churn from one operator to another in search ofa lower tariff, differentiation has become a primarystrategic objective. So although the revenueopportunity that Mobile Money presents is huge,mobile operators are increasingly focused on MobileMoneys potential to strengthen their relationshipwith mobile users, giving them a compelling reasonnot to churn away to a lower-priced operator.

Annual Report 2011


30/96

2526

What are the activities that need tobe perormed to oer Mobile Money?For each, is a bank or an operator bestequipped to perorm it?

A basic tenet of corporate strategy is thatcompanies should seek to perform the activitieswhich they are uniquely well-positioned toperformand outsource those which they arentto rms with greater expertise. Banks and operatorsdo this all the time in their core businesses; both,for example, routinely contract with other rmsto handle their security needs. Bharti Airtel hasbecome famous for taking this principle to its logicalextreme; in the so-called Indian model of mobiletelecommunications, Bharti outsources networkinfrastructure, call centres, and retail stores to otherrms, allowing it to focus on understanding andmeeting customer needs.

This logic offers a useful framework for thinkingabout how banks and operators might work togetherto offer Mobile Money. Offering Mobile Money, likeany other product or service, requires carrying outa coordinated set of activities. These are sometimescollectively called the value chain. The diagrambelow, although not exhaustive, lists the importantparts of the Mobile Money value chain.8 Primaryactivities are those which create and deliver theMobile Money service to customers; support activitiesare required in order to carry out primary activities.

How sophisticated services t into the Mobile Moneyvalue chain

Sophisticated oerings like savings, credit, and insurancecan be, and increasingly are becoming, a part o the MobileMoney value chain. M-KESHO, which is described in theappendix, gives customers access to savings, loans, andinsurance. And in other markets, sophisticated services arebecoming part o the Mobile Money value chain in moremodest ways. Banks have begun to engage with the businessowners o Mobile Money services in order to allow theircustomers to move money into and out o mobile walletsrom and to their bank accounts. Airtel Arica (ormerly Zain),or example, makes it easy or banks to integrate securely

with the Zap platorm in order to oer this unctionality tocustomers. In eect, these institutions are helping to create,and becoming part o, an enlarged Mobile Money valuechain that oers users a broader array o services thanpayments alone.

Because oering such services requires the participation oappropriately regulated nancial institutions, the expansiono the Mobile Money value chain in this way is likely toincrease the number o relationships we observe betweenoperators and nancial institutions in the uture.

Marketing Cash-in/cash-out network Technology Customercare

Product and business development

Float holding

License acquisition, regulatory engagement, and compliance

Communications

Branding Liquidity management(superagency)

Access to the handset

Transactional platform

Primary activities

Supportactivities


8 See Competitive Advantage: Creatingand Sustaining Superior Performance byMichael Porter. Strictly speaking, Porter callsvalue chains that are spread across multiplecompanies, like those we are discussing in thispaper, value systems.


31/96

Chapter 2

Banks are uniquely well positioned to performsome of these activities; operators are to perform

others; and third-parties yet others. To assess whichentity is best positioned to take on these services,it is useful to review the generic strengths that banksand operators generally bring to the table. From ourconversations with banks and mobile operators inthe developing world, the picture below emerged.

In the following pages, we attempt to map theassets and competencies of mobile operators

and banks to the segments of the Mobile Moneyvalue chain. Since our scope is broadoperatorsand banks in the developing worldthe discussionbelow necessarily requires making somegeneralisations that wont apply in every market.So at the end of this article, we offer a guide foroperators and banks seeking to structureorrestructurerelationships with each other toconducting this analysis on a local level.

Tangible assets

Intangible assets

Competencies

Banks

Full suite o fnancial services, includingcredit and savings

Deposit-taking license

Branches with trained sta, security,and deep pools o liquidity

ATMs/cash machines

Integration with broaderfnancial system

Secure core banking platorm

Reputation or stability and security

Relationship with fnancial regulator

Risk management, raud deterrence,and regulatory compliance

Retail operations, including liquiditymanagement

Financial product development

Mobile Network Operators

Large and growing customer base,a signifcant proportion o whomare unbanked

Pervasive airtime distribution network

Control over the SIM card on and datachannel to customers handsets

Robust high-volume, low-value transactionprocessing platorm

High mass-market awareness

Trust o consumers as a transaction partner

Relationship withtelecommunications regulator

Building and managing a third-partydistribution network

Mass market brand-building and advertising

Rapid value-added service development

The assets and competencies, relevant to mobile

money, o banks and mobile operators

Annual Report 2011


32/96

2728

Marketing

Branding

The broad level of awareness that already existsof mobile operator brands is a key asset whenit comes to marketing Mobile Money. Equallyvaluable is the accessibility of most operatorbrands: high rates of mobile penetration in mostof the world means that wide swathes of the low-income segment have existing relationships withone or more mobile operators. Indeed, since inthe pre-paid model such relationships require theconsumer to trust their mobile operator to storevalue that they load into their airtime balance untilthey use it, consumers of mobile telephony cometo trust the operators they use regularly.

Banks typically have lower levels of brandawareness, since they compete for the business ofa narrower range of consumers and conne theirbrand-building activities accordingly. Nevertheless,their brands have certain attributes that could behelpful when promoting Mobile Moneyin somecases, a reputation for stability and security. Butthese brand features can be overshadowed byassociations that prove to be liabilities whenmarketing Mobile Money. Many banks in thedeveloping world are perceived as exclusive (notfor people like me) by the poor; this perceptionis sometimes encouraged deliberately when bankscultivate an upmarket image in order to appeal to

the aspirations of potential clients. Such a strategymakes obvious sense when it comes to buildingmarket share in more afuent segments, butposes complications when it comes to marketingMobile Money.

Not all banks nd themselves in this position;a few (like Equity Bank in Kenya, GrameenBank in Bangladesh, and Bank BRI in Indonesia)explicitly target the low-income consumer andhave developed a brand prole to match, andsuch institutions would be well positioned to treatMobile Money as a brand extension. Nevertheless,in practice there are few bank-branded Mobile

Money deployments that target the base of thepyramid. Even WIZZIT and WING, Mobile Moneyservices offered by subsidiaries of banks (in SouthAfrica and Cambodia, respectively), have chosento build new brands from scratch rather than goto market under the brand of their parent banks.The only exceptions to this pattern we know of areStandard Bank Community Banking in South Africaand Zanacos Xapit in Zambia.

Occasionally Mobile Money services are co-branded;

marketing materials for Vodacom M-PESA in SouthAfrica, for example, carry the Nedbank logo. Thisapproach is usually only adopted when requiredby a regulator, probably because co-brandinghas certain costs. First, it can be confusing to thecustomer; second, it exposes both brands to thefortunes of the other. On the other hand, it mayenhance the offerings credibility with customers.

Communications

Operators experience in building and maintaininga mass-market brand and in investing heavilyin mass- market advertising situates them

naturally to take on the responsibility of marketingMobile Money. One of the main reasons TameerMicronance Banks shareholders (a deposit-takingnancial institution in Pakistan) agreed to sellTelenor shares in the bank was Telenors willingnessto invest far bigger sums in marketing easypaisathan Tameer would have been willing or able todo on its own and its expertise in addressing thatconsumer. The subset of banks that have chosento build mass-market brands of their own mightequally have the experience and muscle to marketMobile Money.

At the same time, the challenge of marketing

Mobile Money is signicantly different fromthe task of marketing operators core offering,airtime. Operators who take on this challengend themselves forced to develop new kinds ofcommunications materials (and indeed to adapttheir entire marketing mix) in order to buildawareness of a nancial service, educate customersabout it, and generate demand objectives thatdiffer signicantly from those of a campaign todrive sales of airtime, which the target marketalready knows, understands, and demands.Banks, although usually targeting a narrowersocio-economic band that mobile operators, arelikely to have a better understanding of this kind

of marketing challenge.9

Co-marketing is extremely rare; M-KESHO is theonly example we know of. This is largely becauseof coordination issues: Safaricom and Equity Bankboth note that it took longer to develop marketingcommunications because two design teams wereinvolved, and two sets of approvals were required.(See the appendix for more about the Safaricom-Equity Bank partnership that led to M-KESHO.)

9 See Driving Customer Usage o MobileMoney or the Unbanked (Chapter 3).



33/96

Chapter 2

Cash-in/cash-out network

As discussed previously, the use of a distributedagent network is what transforms the economicsof offering nancial services to the poor from ahigh-xed-cost business to a low-variable-costbusiness. As such it is the lynchpin of a sustainableMobile Money service.

Operators bring experience from airtimedistribution that is relevant to building a networkof Mobile Money agents. Every mobile operator inthe developing world has developed a sophisticatedairtime supply chain that involves a large numberof independent airtime dealers. (bharti airtel airtimeis available for purchase at more than 1.5 million

retailers in India, for example.) Appointing andmanaging channel intermediaries, performingmargin analyses, devising trade promotions, andnding ways to get branded collateral to the fartherreaches of their markets are some of the capabilitiesthat operators have built for distributing airtimeand that can be leveraged when building a MobileMoney agent network.

Not having traditionally relied on independentagents, most banks lack this expertise. However,an interesting regional exception to this rule isLatin America, where banks in countries like Peruand Brazil have built networks of banking agents

themselves. Although these are smaller than airtimedistribution networks, they have developed know-how around training and monitoring agents whoperform transactions and collect KYC information.This represents both a capability and an asset:these banks have the know-how to create an agentnetwork and they can leverage the network theyvealready built to serve as cash- in/cash-out points fora Mobile Money service.

Finally, in Latin America and in India, a specialkind of third party has emerged, sometimes dubbedan agent aggregator, which takes on the task ofbuilding and managing a network of agents on

behalf of the business owner of a Mobile Moneyservice. These players recognise the value ofsuch agent networks (and, in some cases, realiseeconomies of scale by selling the use of itsagents to multiple Mobile Money or other serviceproviders), although their ability to build andmaintain them will vary.

Liquidity management (superagency)

Despite rarely having experience building agentnetworks themselves, there is one component ofagent network management where banks can addsignicant value to a Mobile Money deploymentwithout building any new capabilities: liquiditymanagement. One of the biggest challenges facingMobile Money services is the need to keep agents,particularly in rural parts of the country wherecustomers primarily seek to perform cash-outtransactions, stocked with enough cash to meetdemand. Banks have established cash logisticsnetworks and instituted appropriate securitymeasures to maintain deep pools of liquidity intheir branches; as such, these branches can support

the agent network by allowing agents to exchangeelectronic value for cash in their branches. Bankswhich play this role, sometimes called superagents,are usually compensated with a per-transaction feethat may increase with the size of the transaction;the fee can be charged to the agent, the masteragent,or the operator.10

What makes a bank an attractive superagent?Primarily, a large branch network in rural areas,where agents are most desperate for cash. Ruralbanks and other banks that target the poor are mostlikely to have such networks.

Why would a bank want to serve as a superagent?Serving as a superagent offers banks an additionalrevenue streamone that is particularly attractivein branches that suffer from low capacity utilisation.(If transaction values are signicant, however, theywill force the bank to assume new costs: not just forheadcount, but also to move cash where it is neededwithin the branch network.) Serving as a superagentalso allows banks that have clients that are agents ofthe Mobile Money service provide those clients withthe convenience of being able to rebalance their oatat the same time they perform banking transactions.

Occasionally, banks offer to use their branches not

as superagents that serve agents, but as agents thatserve customers. United Bank for Africa (UBA) hasrecently forged an agreement with MTN to servethis function in Uganda, hoping not only to earntransaction fees but also cross-sell users on full UBAbank accounts and other products once theyre inthe branch. However, it remains to be seen whetherthis model is sustainable in the long term, given thehigh xed costs of formal bank branches discussedin the introduction to this article.

10 For more on superagents, masteragents,and liquidity management, see Building,Managing, and Incentivising a Network oMobile Money Agents by Paul Leishman andNeil Davidson (http://mmublog.org/global/gsma-publish-2010-mobile-money-or-the-unbanked-annual-report-2/).

Annual Report 2011


34/96

2930

Finally, it is worth highlighting an important assetthat banks can leverage in a Mobile Money service:its network of ATMs (i.e., cash machines). ATMscan complement a network of independent agentsas an option for customers seeking to cash out.Making use of this asset typically requires eitherissuing users ATM cards that are linked to theirMobile Money account (the option adopted bySMART Money in the Philippines) or undertakinga software upgrade to ATMs to allow customersto initiate and authenticate a withdrawal with theirPIN rather than with a card (the route takenby M-PESA in Kenya).

Technology

Transactional platorm

Mobile Money services require the developmentand maintenance of a transactional platform thatcreates individual accounts (mobile wallets)for customers and agents; processes movementsof value between accounts; and interfaces withhandsets, billers, and the core mobile platform.

Both mobile operators and banks have extensiveexperience operating transactional platforms,although they bring complementary strengths tothe table: banks stress the importance of integrity

and robustness when it comes to core bankingsystems, while operators rst priority for theirairtime billing platforms is stability and speedwhen handling huge volumes of transactions.

In practice, however, banks and operators rarelybuild their own Mobile Money transactionplatforms, because there are a host of third-partyproviders offering them in the marketplace.The role of the bank or the operator is usuallytherefore conned to selecting the vendor,providing business rules and other specications,developing APIs for systems integration, and (inmany cases) hosting and operating the platform.11

Given the complementary standards by which banksand operators evaluate transactional platforms,operators can consult with their bank (or vice versa)when selecting a technology solution to be surethat it meets the needs of each participant in thevalue chain.

Access to the handset

To offer users of a Mobile Money service the abilityto initiate transactions on their handset, a datachannel and user interface must be established.Generally speaking, it is difcult to offer customersa user-friendly experience without the mobile operatoreither (1) embedding a menu for the Mobile Moneyservice on the SIM card or (2) assigning a USSD shortcode and providing access to the USSD gateway.12

This is the single asset necessary for Mobile Moneywhich banks are unable to build on their own.But banks can negotiate with operators for accessto the handset in either of these two ways describedabove. In South Africa, SIM cards of all the major

mobile networks carry mobile banking applications,allowing users to access their existing bank accountsfrom the handset, while WIZZIT has secured accessto the USSD channel from the three leading operatorsin the market.

Customer care

Both mobile operators and banks run, or outsource,call centres that cater to their existing customers.In principle, then, either is well-positioned to setup this function for a Mobile Money serviceparticularly since both banks and operators havefound that they train a sub-set of their call-centrestaff to deal with Mobile Money inquiries in order

to effectively resolve problems for customers.Similarly, banks and most operators haveexperience running walk-in customer care points(branches in the case of banks, and agship storesor customer care centres in the case of mobilenetwork operators).

Float holding

As noted in the introduction, oat is always heldby a bank and never by a mobile network operator,because only banks are licensed to take deposits.

Why would a bank want to hold oat for a

Mobile Money service?First, banks make money on deposits by chargingborrowers higher interest rates than they paydepositors, and they can make money on oatholdings in exactly the same way.13 If a MobileMoney service achieves signicant scale, this canbecome a very large deposit. And its an unusuallystable deposit: because it represents the holdingsof many end users and agents, it is unlikely touctuate in value signicantly over time. Second,banks can charge mobile operators transaction fees.11 The acronym API stands or application

programming interace, which reers to theinterace o one piece o sotware (in thiscase, a payments platorm) that allows it tointerace with another.

12 It is technically possible or banks to oerMobile Money without working with operatorsby using ully open mobile channels. But eacho these poses signifcant challenges or costs:SMS interaces are difcult to use and are notsecure; voice is expensive; and the mobile webis inaccessible to most low-income customers.

13 Although in some markets, banks struggleto place their holdings in this way, limiting thevalue that they capture rom deposits alone.



35/96

Chapter 2

Since oat accounts can be high-transaction-volumeaccounts, these fees can be considerable. Third,there is at least one indirect benet of holding oat.Clients of a bank holding oat for a Mobile Moneyservice are sometimes able to convert depositsin their own accounts into e-money more quicklythan others, because an intrabank transfer is fasterthan an interbank one. Some M-PESA agentshave opened accounts at CBA to take advantageof this difference.

Operators need not use only one bank to hold oat.In

Date post:	06-Apr-2018
Category:	Documents
Upload:	shazins
View:	217 times
Download:	0 times

Mobile Money Annual Report 2011

Documents