Agent Network Accelerator Survey:
Uganda Country Report 2013
January, 2014
Contributing Authors: Kimathi Githachuri, Mike McCaffrey, Leena Anthony
Annabel Lee, Anne Marie van Swinderen, Graham A. N. Wright
1
Project Description
Through the financial support of the Bill & Melinda Gates Foundation, MicroSave is
conducting a four-year research project in the following eight focus countries as part of
the Agent Network Accelerator (ANA) Project:
Africa
Kenya
Nigeria
Tanzania
Uganda
Bangladesh
India
Indonesia
Pakistan
Asia
Research findings are disseminated through The Helix Institute of Digital Finance.
Helix is a world-class institution providing operational training for digital finance
practitioners.
2
Focus of Research
The research focuses on operational determinants of success in agent
network management, specifically:
Quality of
Provider
Support
Agent &
Agency
Demographics
Core Agency
Operations
Liquidity
Management
Business
Model Viability
3
The Research Is Based On 2,028 Nationally Representative Agent Interviews
Data
collection
occurred
in
June/July
2013,
using a random
route methodology
based
on
the
displayed
agent
census.
Achieved Sample
373, 18%
540, 27%
Red points represent a
census
of
agents
conducted by Brand
Fusion in 2013.
Blue ones are the ones
interviewed
for
this
research.
1115, 55%
Kampala
Non-Kampala Urban
Rural
Sample Profile*
Key
Providers***
A
B
C
Location
Kampala
NonKampala
Urban
322
102
66
838
208
167
Exclusivity
Rural
392
96
117
Exclusive
1,364**
148
134
NonExclusive
188
258
216
Dedication
Dedicated
NonDedicated
713
170
137
*Note this table shows results only for the top three providers. It sums to 2,308 as it represents all providers served by agents. i.e.
if an interview was done with an agent serving three providers, it is counted three times in this table.
** Note the high concentration of Provider A exclusive agents (1,364) has significant impact on the overall findings.
***Provider names have been anonymized to maintain confidentiality.
839
236
213
4
Uganda Overview
The quantity of agents in Uganda is expanding rapidly and profits are
high; however, the next phase of development needs to focus on the
quality of service provided.
In general agents are satisfied with their profits, rebalance
fairly easily and are dedicated to remaining as agents.
However, agents primarily face challenges on fraud and
security, communication with their provider, and system downtime
5
Providers’ Market Share Of National Agent Network
Warid
17%
Airtel
14%
MTN
63%
Kampala City AMS
Ezee Money is the first third-party provider. It
offers ALL mobile money services and
aggregates them using terminals
Non-Kampala Urban AMS
Others
Warid 6%
Others
Warid 7%
19%
Airtel
13%
During this research period, Airtel and Warid
announced their acquisition and merger plans
Others
6%
Others
Warid 6%
16%
15%
MTN
61%
Airtel
17%
Rural AMS
MTN
62%
Airtel
13%
MTN
65%
Agent market share is defined as the proportion of cash-in/cash-out (CICO) agents by provider.
6
Daily Transaction Levels* Show Healthy Business For Agents
Total Transactions Per Day
Percent of Respondents
30%
Median Transactions Per
Day
The most common band is
21 to 30 transactions a day
25%
Kenya
Tanzania
46
31
Uganda
30
20%
50% of agents are now doing at
least 30 transactions a day
15%
10%
This value shows
growing market
maturity
5%
%
130+
121-130
Kampala
Non-Kampala Urban
Rural
Total
* Numbers represent transactions per day by selected provider, not overall volumes for the agency.
111-120
101-110
91-100
81-90
71-80
61-70
51-60
41-50
31-40
21-30
11-20
1-10
Number of Transactions
7
Largest Stated Barriers To Daily Transactions*
6.00
5.00
Many locations in Uganda
seem saturated
This represents
opportunities for product
development and marketing
communication
Rank
4.00
3.00
2.00
1.00
0.00
Too many other
Lack of
Individual
Lack of
Too often have
Doing more
Too busy to do
agents
resources to buy clients demand awareness of only either cash business means
anymore
competing for
enough float for service is not service among or e-float when too much more
business
business
very regular
potential
the client is
risk of fraud or (already have
customers in the asking for other
robbery
lines of clients)
area
* These scores are weighed averages of rankings, so that higher scores represent dimensions receiving a higher ranking.
8
The Lack Of Offerings Means Potential For Product Innovation
Products And Services Offered In The Country
120%
Percentage of Respondents
100%
Agents report the process takes too long, and many are not even sure what the
commission earned is.
This indicates a very
99%
100%
high level of agentassisted OTC
transactions processed
from the agents’
phone (direct
deposits).
80%
60%
40%
33%
30%
17%
20%
Banking services
(credit, savings,
insurance)
are practically
absent.
17%
1%
Welfare/Social
Savings deposits to a bank
Insurance
Products and Services
Credit
Airtime top-up
Bill payments
Money transfer
Cash-out (withdrawals)
Cash-in (deposit)
Account opening
%
9
The Vast Majority Of Agents Are Profitable*
Profit Per Month
35%
Percent of Respondents
30%
Median Monthly Profit
($US)
11% of agents are making a loss.
24.1%
23.0%
25%
40% of agents are making at least
US$100 of profits a month
20%
Kampala
Non-Kampala
Urban
Rural
86
Total
78
11.0%
10%
8.4%
5%
4.0%4.0%
2.3%2.3%
2.5%
0%
74
The high profitability
in rural areas is a
combination of health
revenues with low
OpEx.
14.8%
15%
78
.4% .4% .8% .6% .2% .0% .6% .0% .0% .2% .0% .2%
950.01+
900.01-950
850.01-900
800.01-850
750.01-800
700.01-750
650.01-700
600.01-650
550.01-600
500.01-550
450.01-500
400.01-450
350.01-400
300.01-350
250.01-300
200.01-250
150.01-200
100.01-150
51-100
1-50
Breaking even
Making losses
Profit in (US$)
Kampala
Non-Kampala Urban
Rural
Total
* Is calculated by subtracting expenses from total earnings from all providers served.
10
Risk of Fraud Most Burdensome To Agents’ Business*
6
Biggest Agent Management Issues
First and third most prevalent issues
are security related, with the risk of
fraud being the biggest concern
5
Ranks
4
Although customer
service does not
often make the
priority list for
providers, it
certainly does for
agents
3
2
1
0
Risk of fraud
Dealing with Threat of Armed Not making
Time spent
Time spent on
Time spent in
customer service
robbery
enough money to
teaching
float
training from
when something
cover costs
customers about management service provider
goes wrong
the product
11
* These scores are weighed averages of rankings, so that higher scores represent dimensions receiving a higher ranking.
Comparison Of Median Profits* By Existing Dimensions
100
90
89
Median Profit (US$)
80
This is an interesting
finding considering a
majority of agents are
female
The absence of differences along these attributes is
intriguing. For Exclusivity specifically, it seems that the
highest performing agents stay exclusive, balancing the fact
that non-exclusive agents receive commissions from more
than one source.
78
78
78
78
Dedicated
Non-Dedicated
Exclusive
Non-Exclusive
70
53%
60
58
50
40
30
20
10
0
Male
Female
* Is calculated by subtracting expenses from total earnings from all providers served.
12
Median Profits* Lowest In the First Year of Operations
140
Years Of Operations & Profitability
We expect the rate of
increase to taper over
time as the market
matures
120
Median Profit (in US$)
100
80
Agencies in their
second year of
operation can expect to
make 81% more profits
78
117
117
50%
60
43
40
81%
20
0
Less Than One Year
1
2
* Is calculated by subtracting expenses from total earnings from all providers served.
3 Or More Years
13
More Than Half Of Agencies Have Been Operating For Less Than
One Year
Years Of Operation Of Agencies
The amount of new agencies
was quite surprising and
could represent robust
growth and/or a high agency
churn rate
3 or more years,
9%
2 years, 12%
Less than 1 year,
52%
1 year, 27%
Given 78% of agents
interviewed predicted they
would continue with the
business next year, it seems
more likely that the majority
of this is growth
14
Median Monthly Revenue Is Well Above GNI Per Capita
Median Revenue (US$)
Median Revenue In US$*
180
160
156
140
Agencies in urban areas
outside Kampala are
earning 16% lower than
the country average.
136
136
117
120
Monthly
GNI per
capita =
US$95
100
80
60
40
20
0
Kampala
Non-Kampala urban
Rural
Total
*Revenue reported here pertains to all providers being served. Therefore for non-exclusive agencies, their total
revenue is reported here with regards to all the providers they serve.
15
Agents Report Diverse Operating Costs, From Almost Nothing To
Well Over 100 US$ A Month
30.0%
Operational Expense Per Month
Operational costs are mainly
driven by rental, wages and
rebalancing costs
25.0%
Percent of Respondents
Median:
Kampala:
Non-Kampala Urban:
Rural:
Total :
78 US$
58 US$
52 US$
58 US$
20.0%
20% of agents in Kampala are
paying US$ 100 or more in
monthly operational expenses
15.0%
10.0%
5.0%
.0%
0 - 19
20 - 39
40 - 59
60 - 79
Kampala
80 - 99 100 - 119 120 - 139 140 - 159 160 - 179 180 - 199
Non-Kampala Urban
Rural
Total
200+
16
Surprisingly Many Agents Have Balanced Needs For E-Float & Cash
Regions:
Central
About
The
Same,
44%
Cash,
24%
EFloat,
32%
Northern
About
The
Same,
35%
Cash,
7%
EFloat,
57%
Eastern
About
The
Same,
50%
Cash,
19%
E-Float,
31%
Western
About
The
Same,
49%
Cash, 17%
Except for the Northern region
the majority of agents in each
region are reporting a balanced
need for e-float and cash
This may represent a general
misunderstanding of agent
needs, as many providers
assume that predominantly
rural regions, such as the North,
have more need for cash than efloat.
E-Float,
34%
17
Rebalancing Is Easy For Agents In Terms Of Time And Money
Time Taken To Nearest Rebalance Point (Minutes)
Kampala
Non-Kampala
Urban
Rural
National
Average
< 5.00
23%
34%
22%
29%
5.00-14.99
39%
46%
40%
43%
15.00-24.99
24%
11%
22%
16%
25.00-34.99
13%
5%
10%
8%
35.00-44.99
0%
2%
2%
2%
45.00-54.99
1%
1%
2%
2%
55.00+
0%
0%
2%
1%
Non-Kampala Urban agents have
more convenient rebalancing
points relative to the national
average
72% of agents spend less
than 15 minutes
travelling to their nearest
rebalancing point
9% of agents responses
were not included here
as they reported they
did not travel to
rebalance.
Agents tend to pay little or nothing to rebalance: 66% have costs of less than US$1
18
Frequency Of Monthly Withdrawals Seems Low
Percent of Respondents
The majority of agents
withdraw cash five times or
Monthly Cash Withdrawals For Rebalancing
18% less per month.
16%
14%
12%
A significant
minority
rebalances daily
Monthly cash withdrawal
behaviour seems to be similar
across all geographies
10%
8%
6%
4%
2%
%
0
1
2
3
4
5
6
7
8
9
10 11 12 13 15 16 17 18 20 21 22 23 25 26 27 28 30
Number of Withdrawals
Kampala
Non-Kampala Urban
Rural
19
Three Transactions Are Denied Each Day Due To Lack Of Float
Transactions Denied Due To Unavailability Of Float
30%
Percent of Respondents
27%
25%
Only 27% operate
consistent and optimal
levels of float
Median:
This is equivalent
to 10% of average
daily transactions
being lost across
the country
Kampala:
Non-Kampala Urban:
Rural:
3
3
3
National Median:
3
20%
16%
15%
15%
11%
35% of agents lose at least five or more
transactions per day due to lack of float
10%
7%
5%
5%
5%
3%
1%
3%
%
%
0
1
2
3
4
5
6
7
6%
8
Number of Transactions Denied Per Day
9
10
More
Than
10
20
Unpredictable Demand Seen As Greatest Impediment To Float
Management*
Much of the predictable
behaviour is not effectively
This is an indication that
communicated by providers to
the quality of agents
6.00
agents
recruited have inadequate
resources (poor selection)
5.00
Rank
4.00
This may be indicative of
the challenges providers
face in master agent
recruitment and training
3.00
2.00
1.00
0.00
Unpredictable
fluctuations in
client demand
Lack resources in Have to shut store The cost incurred
Time taken at
Rebalance points
Travel time to
general to buy a
to go get more is too much to do it rebalance point is often do not have rebalance point is
sufficient amount
float
frequently
too long
cash/float
too long
that will last
* These scores are weighed averages of rankings, so that higher scores represent dimensions receiving a higher ranking.
21
Improving The Quality Of Agent Support Presents A Large
Opportunity For Providers
Training
94% report receiving training
 46% from a provider
 13% from master agent
 45% from an employer
57% of agents have never undergone refresher
training
Operational Support:
Only 33% of agents were visited by the
provider, whereas 46% of agents report not
being visited at all. Of those who were visited
35% report they were with no fixed frequency.
Call Centre:
93% of agents were aware of a call centre and rated it a 4.6 out of
7 in terms of its ability to resolve its issues.
22
Recurrent Service Downtime Is Affecting Transaction Levels
Unreliable service is a challenge for most agents:
92% of agents report having experienced downtime in
the past.
Huge differences in providers, with some performing
considerably better than others
Only 48% of agents report receiving prior warning
about downtime, however two thirds report that
information given is inaccurate
There is a great reported variation per provider on
amount of transactions lost per occurrence of server
downtime with the median reported to be 10
transactions per day.
23
Outstanding Attributes Of Agent Network Management
Uganda is one of the most developed digital finance
countries in the world with a robust agent network:
Daily transaction volumes are high from Kampala
through to the rural areas surveyed
Healthy profits and revenues
Strong competition and evidence of market
saturation in some areas
Cost-effective rebalancing systems
Most agents intend to continue with the business
24
Opportunities For Improvement
Agents are struggling with a number of issues. Here are some
suggested areas of focus:
Large challenges on fraud and other security issues
Lack of product diversity especially on the agent-level as the
large majority of transactions are CICO
Service down-time is greatly limiting number of transactions
per day
Customer service support is not adequately attended to
Agents encounter frequent direct deposits (agent-assisted
OTC transactions) as shown by the money transfer incidence
High incidents of newly operating agencies needing close
support that is not forthcoming
Most agents close to rebalancing points resulting in limited
geographical reach and financial inclusion
Agent selection seems to be a significant challenge and
barrier to increased transactions
25
Thank You
www.helix-institute.com
[email protected]
Helix Institute of Digital Finance
Helix Institute
26