Quantifying the telecommunication
opportunity at the base of the pyramid in
South Africa: A retail perspective
I Meyer
23228687
Mini-dissertation submitted in partial
fulfilment of the
requirements for the degree Magister
in
Business
Administration at the Potchefstroom Campus of the North-West
University
Supervisor:
Prof RA Lotriet
i
ACKNOWLEDGEMENTS
My thanks and appreciation to the following:
To our Lord for this amazing opportunity.
To my family and friends who encouraged me to start my MBA journey and
supported me throughout the three years.
To my class mates who I had the privilege to study with. It’s been long but
well worth it. A special thanks to the Wasserval group members, D. Egberts,
P. Goosen, C. Beyers, O. Jansen van Vuuren, P. Steyvesant, C. Fransman
and S. de Beer for keeping the balance.
To my study leader, Prof. R. Lotriet for his insights and attention to detail to
ensure a quality research report.
ii
ABSTRACT
This study focuses on the telecommunication industry, specifically the mobile phone
market at the base of the pyramid (BOP). A supply vs. demand stance is taken
whereby demographic data offer insight into the demand while the location of
telecommunication retail stores constitutes supply. The study furthermore makes
extensive use of a GIS (geographical information system) which offers deeper insight
into data and different applications thereof. Given the extensive nature of the data
used in assessing the national market, a dashboard was developed as part of this
research to ease data interpretation. The online map (GIS) and dashboard form an
integral part of this report.
Literature supports the targeting of the BOP as a viable market given the high
volume of people in this market segment. Unconventional methods are, however,
required to sustainably cater to this market. The development of multiple channels to
target potential consumers has resulted in a dilution of the market in the retail
environment. The retail channel however remains important in any company’s
strategy to target the BOP. Telecommunication companies can not only benefit from
the BOP but also offer benefits to the BOP. The World Bank has reported figures
that show a 0.8% increase in GDP for every 10% increase in mobile penetration.
Different sources identify the BOP by different income ranges. It became evident,
however, that internationally the average applied to identify the BOP is households
earning less than USD 3,000 (ZAR 31,440 at an exchange rate of R10.48 / USD) per
annum. Although somewhat higher than the international average, the available data
dictated that South Africa’s BOP be identified as households earning less than
R38,200 per annum for the purposes of this research. Data indicate that 89% of
households in South Africa have a mobile phone. By comparing the ownership of
existing household goods this research found that of the 11% of households not
owning a mobile phone, 6% would be willing to adopt a mobile phone.
A tiered approach is followed in assessing the telecommunication opportunity for
mobile phones in the BOP. The first tier assesses the entire market (all households
iii
in South Africa) at a municipal level. By including the total market, the opportunity in
the BOP is put into perspective. The result was that the total BOP market offers a
potential market of R563 million per month through 13.7 simcards. At the other end
of the economic pyramid, the ROP offers a market of R2416 million per month
through 14.6 million simcards. Thus, even though the BOP offers 48% of the total
volume in the market, the value is only 19% of the total market. From a coverage
perspective, 42% of BOP households are not covered by a telecom retailer whilst
only 27% of the ROP households are not covered. A market of R247 million (through
5.8 million simcards) has been estimated in the BOP opposed to a R379 million
market (through 5.8 million simcards) in the ROP.
The second tier makes use of a case study to determine the viability of targeting the
BOP. Moruleng Mall’s catchment area was analysed within the Moses Kotane
municipality that offered a high opportunity as determined in the first tier of analysis.
This case study made use of gravity modelling and found that Rustenburg’s retail
offering would have limited influence and as such telecom retailers would have to
revisit their strategy for the area. A number of shopping centre developments in rural
areas were highlighted. Effectively while these developments are taking retail closer
to the BOP or rual population, the market is diluted. From a retailer perspective, this
makes it difficult to target an entire area through presence in one specific retail node
or town.
The ultimate finding of this this research suggests that it is in fact possible to target
the BOP
– however, that it is the ROP located between the BOP households that
makes this a viable market. This suggests that it is rather not a question of the
viability in targeting the BOP specifically but targeting the more dense rural areas
that offer opportunity.
KEYWORDS
Geo-demographics, Retail, BOP, Mobile Phone, Technology adoption, Location
Analytics, GIS, Retail Strategy
iv
OPSOMMING
Hierdie studie fokus op die telekommunikasie-industrie, meer spesifiek die
selfoonmark wat deel is van die basis van die piramide (BOP). ‘n Aanbod
versus-aanvraag houding word ingeneem, waar demografiese data insig bied in die
aanvraag, terwyl die plasing van telekommunikasie-kleinhandel die aanbod uitmaak.
Die studie maak ook uitgebreid gebruik van ‘n GIS (Geografiese Inligtingstelsel) wat
meer insig bied in die data en die verskillende toepassings van die data. Gegewe die
uitgebreide aard van die data wat gebruik word in die nasionale mark is ‘n
“dashboard” ontwikkel wat meer insig bied in die data en die gepaardgaande
toepassings om die interpretasie van die data makliker te maak. Die aanlynkaart
(die GIS) en die “dashboard” maak ‘n integrale deel uit van hierdie verslag.
Die literatuur ondersteun die teikening van die BOP as ‘n lewensvatbare mark
gegewe die groot aantal mense in hierdie marksegment. Nie-konvensionele
metodes word egter benodig om volhoubare voorsiening te maak vir hierdie mark.
Die ontwikkeling van veelvuldige kanale om potensiële verbruikers te teiken het
uitgeloop op ‘n verdunning van die mark in die kleinhandelomgewing. Die
kleinhandelkanaal bly egter belangrik in enige maatskappy se strategie wat gemik is
op die BOP. Telekommunikasiemaatskappye kan nie slegs voordeel trek uit die BOP
nie, maar moet ook voordele bied aan die BOP. Die Wêreldbank het syfers
aangekondig wat ‘n 0.8% toename in GDP aantoon vir elke 10% toename in
selfoonpenetrasie.
Verskillende bronne identifiseer die BOP binne verskillende inkomste-omvange. Dit
het egter duidelik geword dat die internasionale gemiddeld wat op die BOP toegepas
word huishoudings is wat minder as USD 3,000 (ZAR 31,440 teen n wisselkoers van
R10.48 / USD) per jaar verdien. Hoewel dit ietwat hoër is as die internasionale
gemiddeld, het beskikbare data gesuggereer dat Suid-Afrika se BOP geïdentifiseer
moet word as huishoudings wat minder as R38,200 per jaar verdien (vir doeleindes
van hierdie navorsing). Data toon aan dat 89% van huishoudings in Suid-Afrika
selfone het. Deur die eienaarskap van bestaande huishoudelike goedere te vergelyk
is daar in hierdie navorsing bevind dat van die 11% van huishoudings wat nie
selfone besit nie, 6% wel gewillig sal wees om een aan te skaf.
v
‘n
Gelaagde
benadering
is
gevolg
in
die
assessering
van
die
telekommunikasiegeleentheid vir selfone in die BOP. Die eerste laag assesseer die
hele mark (alle huishoudings in Suid-Afrika) op die munisipale vlak. Deur die hele
mark in te sluit word die geleenthede vir die BOP in perspektief geplaas. Die
resultaat was dat die totale BOP-mark ‘n potensiële mark van R563 miljoen per
maand bied deur 13.7 miljoen simkaarte. Aan die ander kant van die ekonomiese
piramide bied die ROP (res van die piramide) ‘n mark van R2416 miljoen deur 14.6
miljoen simkaarte. Dus, hoewel die BOP 48% van die totale volume in die mark bied,
is die waarde slegs 19% van die total mark. Vanuit ‘n dekkingsperspektief is dit so
dat 42% van BOP huishoudings nie gedek word deur ‘n telekomkleinhandelaar nie,
terwyl slegs 27% van die ROP huishoudings nie gedek word nie. Daar is ‘n geskatte
mark van R247 miljoen (deur 5.8 miljoen simkaarte) in die BOP vergeleke met ‘n
R379 miljoen mark (deur 5.8 miljoen simkaarte) in die ROP.
Die tweede laag maak gebruik van ‘n gevallestudie om die geldigheid te bepaal van
die teikening van die BOP. Moruleng Mall se opvanggebied is ontleed (binne die
Moses Kotane-munisipaliteit) aangesien dit ‘n goeie geleentheid gebied het soos
bepaal binne die eerste laag se ontleding. Hierdie gevallestudie het gebruik gemaak
van swaartepunt-modellering en bevind dat Rustenburg se kleinhandelaanbod ‘n
beperkte invloed sou hê en dat telekomkleinhandelaars dus weer hulle strategieë in
die area sou moes bedink. ‘n Aantal inkoopsentrumontwikkelings in plattelandse
gebiede is onder die loep geplaas. In effektiewe terme is dit waar dat hoewel hierdie
ontwikkelinge die kleinhandel nader bring na die BOP of the plattelandse bevolking,
die mark verdun is. Vanuit ‘n kleinhandelperspektief maak dit natuurlik sake moeilik
as mens ‘n hele gebied wil teiken binne een spesiefieke nodus of dorp.
Die uiteindelike bevinding van hierdie navorsing suggereer dat dit inderdaad
moontlik is om die BOP te teiken
– maar dat dit die ROP, tussen die BOP
huishoudings, is wat dit ‘n lewensvatbare mark maak. Dit suggereer dat dit nie
soveel ‘n vraag is oor die teikening van die BOP spesifiek nie, maar teikening van
die digter-bevolkte plattelandse gebiede is wat geleenthede bied.
SLEUTELWOORDE
Geo-demografie, kleinhandel, BOP, selfoon, tegnologie-aanvaarding, plekontleding,
GIS, kleinhandelstrategie
vi
APPLICATIONS OUTSIDE THIS DOCUMENT TO CONSIDER
The nature of this research project required the inclusion of a dashboard and online
mapping solution as an integral part of the research. However all references made to
the dashboard and online map is included in this research through still images.
Some recommendations to consider when viewing the dashboard and online
mapping solution are.
Recommendations for easy viewing and usage of the dashboard.
Copy dashboard to local computer drive;
Microsoft Excel 2010;
Adequate computer processing speed given the size of the model.
After opening the model and selecting the first criteria allow a few seconds for
the model to update. Ultimate ease in using the model will be dependent on
computer specifications.
The dashboard is included on the accompanying CD to this research. A final
dashboard result is included in Annexure I.
Recommendations for easy viewing of the interactive map.
Adequate internet connection speed;
Google Chrome browser;
After clicking the hyperlink (globe) in the dashboard allow a few seconds for
the map to open. After initial opening, navigating on the map should be faster
however will be dependent on internet connection speed.
The map can be accessed at the following address:
https://www.google.com/fusiontables/embedviz?q=select+col2%3E%3E0+from+1ydVxDG60Fry9oAmg
vii
TABLE OF CONTENTS
ACKNOWLEDGEMENTS ... i ABSTRACT... ii KEYWORDS ... iii OPSOMMING ... iv SLEUTELWOORDE ... vAPPLICATIONS OUTSIDE THIS DOCUMENT TO CONSIDER ... vi
LIST OF TABLES ... xi
LIST OF ACRONYMS ... xii
CHAPTER 1: SCOPE AND NATURE OF THE STUDY ... 1
1.1 INTRODUCTION ... 1
1.2 PROBLEM STATEMENT ... 3
1.3 OBJECTIVES OF THE STUDY ... 3
1.3.1 Primary objective ... 3
1.3.2 Secondary objectives... 3
1.4 RESEARCH METHODOLOGY ... 4
1.5 SCOPE OF THE STUDY ... 6
1.6 LIMITATIONS OF THE STUDY ... 6
1.7 CONTRIBUTION OF THE STUDY ... 8
1.8 LAYOUT OF THE STUDY ... 9
1.9 SUMMARY ... 10
CHAPTER 2: THE MOBILE PHONE MARKET IN THE BOP ... 11
2.1 INTRODUCTION ... 11
2.2 BASE OF THE PYRAMID (BOP) ... 13
2.2.1 Identifying the BOP... 14
2.2.2 Evolution of BOP thinking ... 16
2.2.2.1 Public Aid and Charity ... 16
2.2.2.2 BOP as consumers ... 17
2.2.2.3 BOP as producers ... 18
2.2.2.4 BOP as business partner ... 19
2.2.3 Adding context to the BOP opportunity ... 20
viii
2.3 ADOPTION MODELS ... 24
2.3.1 Generic Technology Adoption Models ... 25
2.3.1.1 Technology Acceptance Model ... 26
2.3.1.2 Unified theory of acceptance and use of technology ... 29
2.3.2 Mobile Phone Adoption Model ... 30
2.4 CONNECTING MOBILE PHONE ADOPTION WITH THE BOP ... 31
2.4.1 BOP mobile phone adoption explained by the Diffusion of Innovation ... 32
2.4.2 Social responsibility through telecommunication at the BOP ... 33
2.5 EVIDENCE ON MOBILE PHONE EXPENDITURE ... 35
2.6 MARKET ENTRY ... 38
2.6.1 A Multichannel Offering ... 39
2.6.2 Accessing the BOP ... 41
2.6.3 Current Telecom retail ‘tools’ ... 45
2.7 SUMMARY ... 46
CHAPTER 3: EMPIRICAL INVESTIGATION – THE MOBILE PHONE OPPORTUNITY
AT THE BOP ... 52
3.1 INTRODUCTION ... 52
3.2 GIS IN RETAIL FORECASTING ... 52
3.3 TIERS OF ANALYSIS ... 54
3.4 RESEARCH METHODOLOGY ... 54
3.4.1 The utilisation of various datasets ... 55
3.4.1.1 Boundary data ... 55 3.4.1.2 Demographic data ... 56 3.4.1.3 Contextual data ... 57 3.4.1.4 Retail data ... 57 3.4.1.5 Propensity data ... 59 3.4.2 Software Utilisation ... 60 3.4.3 Forecasting Models ... 60 3.4.3.1 Analogues ... 62 3.4.3.2 Multiple regression ... 62 3.4.3.3 Gravity models ... 63 3.4.4 Summary ... 64
ix
3.5.1 Input data ... 65
3.5.2 Data Modelling ... 66
3.5.2.1 Multiple regression to identify the rate of adoption ... 66
3.5.2.2 Mobile phone consumption and expenditure ... 69
3.5.2.3 Retailers and Shopping Centres ... 70
3.5.2.4 Reworking of input data ... 71
3.5.3 Model Outputs ... 72
3.5.3.1 Results on the model outputs ... 72
3.5.3.2 The development of a telecom dashboard ... 81
3.5.3.3 Spatial view of result ... 94
3.6 THE MOSES KOTANE MUNICIPALITY AS CASE STUDY ... 96
3.6.1 The development of Moruleng Mall ... 98
3.6.2 Methodology: Moruleng second-tier analysis ... 101
3.6.2.1 Data preparation and variable selection ... 101
3.6.2.2 Determining trade areas ... 102
3.6.2.3 SiteMarker result comparison with dashboard output ... 105
3.6.3 Case study findings ... 106
3.7 SUMMARY ... 107
CHAPTER 4: FINDINGS AND RECOMMENDATIONS ... 109
4.1 INTRODUCTION ... 109
4.2 SUMMARY OF THE RESEARCH PROBLEM ... 109
4.3 SUMMARY ON THE BOP LITERATURE ... 110
4.4 MAIN FINDINGS OF THE STUDY ... 111
4.4.1 Suggested model and approach ... 111
4.4.2 The BOP opportunity ... 112
4.4.3 Moruleng Mall as case study ... 114
4.5 STUDY EVALUATION ... 114
4.5.1 Primary objective ... 115
4.5.2 Secondary objectives... 115
4.6 RECOMMENDATIONS FOR FUTURE RESEARCH ... 115
4.7 CONCLUSION ... 116
LIST OF REFERENCES ... 119
x
LIST OF FIGURES
Figure 1: Applications of geographical analysis in retail network planning ... 6
Figure 2: Fixed and Mobile Connections and Penetrations in the Middle East and Africa ... 12
Figure 3: Four tiers of the market ... 15
Figure 4: BOP 1.0 and 2.0 protocols ... 20
Figure 5: The South African Pyramid ... 22
Figure 6: Difference in reporting on annual income ... 23
Figure 7: Household access to the internet... 24
Figure 8: The Original Technology Acceptance Model ... 26
Figure 9: TAM2 - Extension to the original TAM ... 27
Figure 10: TAM3 - Extension to the TAM2 ... 28
Figure 11: UTAUT Research Model ... 29
Figure 12: Model for Mobile Phone adoption ... 31
Figure 13: Diffusion of Innovation ... 33
Figure 14: South African ICT household spending by income segment (in USD) ... 36
Figure 15: A selection of telecommunication companies' EBITDA and monthly ARPU ... 38
Figure 16: Commercial Infrastructure pillars required at the BOP ... 43
Figure 17: New Strategies at the BOP ... 44
Figure 18: Summary of the evolution of BOP thinking ... 47
Figure 19: Influential Variables for the Mobile Adoption Model... 49
Figure 20: Literature contribution aimed at ultimate result ... 51
Figure 21: Inputs to secure competitive advantage through geographical analysis ... 53
Figure 22: Principals used in the main forecasting models ... 61
Figure 23: Multiple Regression Output on household goods ... 67
Figure 24: Households with a mobile phone and willingness to adopt ... 68
Figure 25: Dashboard compilation and workings ... 82
Figure 26: Dashboard page.1, Section 1 ... 83
Figure 27: Dashboard page.1, Section 2 ... 84
Figure 28: Dashboard page.1, Section 3 ... 85
Figure 29: Dashboard page.2 ... 87
Figure 30: Dashboard page.3, Section 1 ... 88
Figure 31: Dashboard page.3, Section 2 ... 89
Figure 32: Dashboard page.3, Section 3 ... 91
Figure 33: Dashboard page.4, Section 1 ... 92
Figure 34: Dashboard page.4, Section 2 ... 92
Figure 35: Dashboard page.4; Section 3 ... 93
Figure 36: Spatial representation of the BOP not covered (through a GIS) ... 95
Figure 37: Moruleng Mall orientation ... 98
Figure 38: Household density map of Moses Kotane municipality and surrounds ... 99
Figure 39: Value density map of the Moses Kotane municipality and surrounds ... 100
xi
LIST OF TABLES
Table 1: Assumptions of the BOP ... 17
Table 2: Changing dominant logic of public policymakers in India with regards to the BOP . 18 Table 3: Difference in Socio-economic Characteristics between Early and Late Adopters .... 32
Table 4: Monthly expenditure on mobile phones ... 37
Table 5: Methodology Overview ... 65
Table 6: Household Mobile Phone Consumption ... 70
Table 7: Number of Telecom Stores by Province ... 73
Table 8: Average direct distance (km) to a telecom retailer ... 73
Table 9: Telecom retailers in the 12 largest shopping centres ... 75
Table 10: Comparison between proximity in different markets ... 75
Table 11: Telecom retailer coverage ... 76
Table 12: Household telecom retailer coverage ... 77
Table 13: Shopping Centre benchmarks ... 78
Table 14: Total Monthly Market... 78
Table 15: Total breakdown of the South African BOP market as per Hammond et al. ... 80
Table 16: Mobile market breakdown by province and part of the pyramid outside coverage 81 Table 17: Top 10 municipalities offering opportunity (not covered by telecom retail) ... 97
Table 18: Attractiveness variables of nodes ... 103
xii