tucan3g business model: case study and field verification

56
ICT-601102 STP TUCAN3G Wireless technologies for isolated rural communities in developing countries based on cellular 3G femtocell deployments D34 TUCAN3G business model: case study and field verification Contractual Date of Delivery to the CEC: 30 th Nov 2015 Actual Date of Delivery to the CEC: 15 th Jun 2016 Author(s): Luis F. Solórzano (EHAS), Ignacio Prieto-Egido (EHAS), Gustavo Ramírez (UCAU), Yury Castillo (CREPIC), César Gómez (CREPIC), Konstantinos Fouskas (KINNO), Vasiliki Gargalianou (KINNO), Paulo Flores (TdP), Diego Ramos (TdP) Participant(s): EHAS, UCAU, CREPIC, KINNO, TdP Workpackage: 3 Est. person months: 3.84 Security: Public Dissemination Level: PU Version: c Total number of pages: 57 Abstract This document describes the TUCAN3G approach to sustainable business for building mobile broadband infrastructure in isolated rural communities of developing countries. Building basic services infrastructure in bottom of the pyramid markets has complexity because we face multiple challenges: social impact, financial sustainability, regulatory framework, appropriate technology, and multiple stakeholders. This complexity requires a socio-technical approach, a global dimension in the engineering solution design. The presented case is in the Peruvian Amazon, but the conclusions can be applied in other similar socio-economic contexts. Keyword list: business model, mobile network infrastructure, isolated rural communities.

Upload: lfsolorzano

Post on 13-Apr-2017

81 views

Category:

Mobile


1 download

TRANSCRIPT

Page 1: TUCAN3G business model: case study and field verification

ICT-601102 STP TUCAN3G

Wireless technologies for isolated rural communities in developing countries based on cellular 3G femtocell deployments

D34

TUCAN3G business model: case study and field verification

Contractual Date of Delivery to the CEC: 30thNov 2015

Actual Date of Delivery to the CEC: 15thJun 2016

Author(s): Luis F. Solórzano (EHAS), Ignacio Prieto-Egido (EHAS), Gustavo Ramírez (UCAU), Yury Castillo (CREPIC), César Gómez (CREPIC), Konstantinos Fouskas (KINNO), Vasiliki Gargalianou (KINNO), Paulo Flores (TdP), Diego Ramos (TdP)

Participant(s): EHAS, UCAU, CREPIC, KINNO, TdP

Workpackage: 3

Est. person months: 3.84

Security: Public

Dissemination Level: PU

Version: c

Total number of pages: 57

Abstract This document describes the TUCAN3G approach to sustainable business for building mobile broadband infrastructure in isolated rural communities of developing countries. Building basic services infrastructure in bottom of the pyramid markets has complexity because we face multiple challenges: social impact, financial sustainability, regulatory framework, appropriate technology, and multiple stakeholders. This complexity requires a socio-technical approach, a global dimension in the engineering solution design. The presented case is in the Peruvian Amazon, but the conclusions can be applied in other similar socio-economic contexts. Keyword list: business model, mobile network infrastructure, isolated rural communities.

Page 2: TUCAN3G business model: case study and field verification

2

Document Revision History DATE ISSUE AUTHOR SUMMARY OF MAIN CHANGES 10 May 2016 a EHAS First complete draft 10 Jun 2016 b EHAS Major revision of structure and content 12 Jun 2016 c URJC Technical revision

Page 3: TUCAN3G business model: case study and field verification

ICT-601102 STP Document number: D34 Title of deliverable: TUCAN3G business model: case study and field verification

TUCAN3G_D34EHASd.docx 3

Executive Summary Living conditions in isolated rural communities are hard because they lack access to the basic services, i.e.: electricity, water, sanitation, health and education. People living in such remote and rural areas of developing countries are usually neglected as a target market, even by public services. There is evidence on the benefits of access to information and telecommunication technologies (ICTs) for spurring economic growth and social inclusion, but the rural-urban digital divide needs to be solved. In the TUCAN3G pilot, the challenge was to solve the mobile coverage gap in isolated rural areas in the Peruvian Amazon. The current solution is to deploy satellite-based rural telephony and internet, but this is very expensive for the majority of users. Satellite links have been the preferred technology for most MNOs as a rural backhaul solution due to the relatively low installation and equipment costs (CAPEX) and short installation time. However, satellite backhaul carries a heavy operating expense (OPEX) which could hinder the profitability and sustainability of the infrastructure in low-income and low-density rural communities. The TUCAN3G project has developed an innovative technical solution for affordable mobile network infrastructures in rural areas of developing countries. The solution proposed will enable to provide mobile services profitably in rural population centres with a few hundreds of users. The TUCAN3G business model is based on a multi-stakeholder partnership between the mobile operator that provides services to end-users, the rural network operator that manages the infrastructure and public/private institutions providing funding to build new tower sites (backhaul network nodes). The most valuable resource provided by the rural network operator are the radio towers; sharing these existing antenna support structures with a MNO for providing new 3G mobile services is the main factor driving cost savings. GSMA and McKinsey estimate that 64% of people without access to mobile coverage, or about 900 million people, live in rural communities. According to ITU’s statistics, more than 2.4 billion people living in rural communities worldwide do not have access to 3G services. Specifically in Peru, with a monthly ARPU of $9, this is a market opportunity for mobile operators estimated at least $216 million annual and a social impact opportunity to improve the life conditions for at least 4 million people. Extending an existing rural network infrastructure requires funding for the new deployments (to build new radio tower sites), and in this context, Public-Private Partnerships can play an important role in fostering long-term investment and economic growth for the rural poor. In this case, the initial capital expenditure (CAPEX) may require a partnership with other public (or private) agents – i.e., Broadband Development National Agencies, Venture Capital Funds, Social Impact Investors, and International Development Agencies – that can subsidize the new deployments. Finally, we present in this document the results of the TUCAN3G business model validation based on a field trial. We measured the traffic of voice and data services provided by TUCAN3G. A field survey was conducted before and after the TUCAN3G pilot. The business case analysis in chapter 3 details the cost and revenue estimation based on the methods and assumptions described in chapter 2. In chapter 4, we present the sensitivity analysis to see what happens if subsidies are not available or how the number of users influences the payback period. The risk analysis, also in chapter 4, is focused on the influence of regulation and public policies on the business model. The market opportunity to scale up the TUCAN3G solution in other site locations where there is previously built radio towers is analysed in chapter 5. Conclusions discuss a multi-stakeholder partnership approach, which is the best business model approach for a new site location. When the radio towers need to be built in the jungle, someone has to provide finance or subsidies. However, subsidies are not needed to build new radio towers in highland rural areas, since the investment costs are much lower and the payback period would be less than 3 years.

Page 4: TUCAN3G business model: case study and field verification

4

DISCLAIMER

The work associated with this report has been carried out in accordance with the highest technical standards and the TUCAN3G partners have endeavoured to achieve the degree of accuracy and reliability appropriate to the work in question. However since the partners have no control over the use to which the information contained within the report is to be put by any other party, any other such party shall be deemed to have satisfied itself as to the suitability and reliability of the information in relation to any particular use, purpose or application.

Under no circumstances will any of the partners, their servants, employees or agents accept any liability whatsoever arising out of any error or inaccuracy contained in this report (or any further consolidation, summary, publication or dissemination of the information contained within this report) and/or the connected work and disclaim all liability for any loss, damage, expenses, claims or infringement of third party rights.

Page 5: TUCAN3G business model: case study and field verification

ICT-601102 STP Document number: D34 Title of deliverable: TUCAN3G business model: case study and field verification

TUCAN3G_D34EHASd.docx 5

Table of Contents 1 INTRODUCTION .............................................................................................................. 11

1.1 Problem statement ......................................................................................................... 11

1.2 Mobile service demand in rural areas ........................................................................... 11

1.3 Appropriate technology solution................................................................................... 17

1.4 Social impact of mobile services .................................................................................. 18

2 METHODS AND ASSUMPTIONS .................................................................................. 20

2.1 Financial metrics ........................................................................................................... 20

2.2 Assumptions ................................................................................................................. 21

2.3 Scope ............................................................................................................................. 23

2.4 Scenarios ....................................................................................................................... 24

2.5 Cost model .................................................................................................................... 26

2.6 Social benefits (SROI) .................................................................................................. 28

2.7 Data sources .................................................................................................................. 29

3 BUSINESS CASE ANALYSIS ......................................................................................... 31

3.1 Financial model ............................................................................................................. 31

3.2 Analysis of results ......................................................................................................... 36

3.3 Non-financial results ..................................................................................................... 37

4 SENSITIVITY AND RISKS ANALYSIS ........................................................................ 39

4.1 Sensitivity analysis ....................................................................................................... 39

4.2 Regulatory issues .......................................................................................................... 44

4.3 Free public services ...................................................................................................... 47

4.4 Other public policies ..................................................................................................... 48

5 SCALING UP THE SOLUTION ...................................................................................... 49

5.1 Leveraging existing infrastructure ................................................................................ 49

5.2 Outsourcing network operation..................................................................................... 53

5.3 Building new infrastructure .......................................................................................... 54

6 CONCLUSIONS ................................................................................................................ 55

Page 6: TUCAN3G business model: case study and field verification

6

References

[Argandoña14] Argandoña, D and More, J. Estimación de torres en las redes móviles para el año 2025 en el Perú. Gerencia de Políticas Regulatorias y Competencia, Subgerencia de Análisis Regulatorio. OSIPTEL 2014.

[Banerjee11] Banerjee, A. V., & Duflo, E. Poor economics: A radical rethinking of the way to fight global poverty. Public Affairs, 2011.

[Bhavnani08] Bhavnani, A., Chiu, R. W. W., Janakiram, S., Silarszky, P., & Bhatia, D. The role of mobile phones in sustainable rural poverty reduction. World Bank ICT Policy Division, 2008.

[Cabral06] Cabral, L., Farrington, J., &Ludi, E. The Millennium Villages Project–a new approach to ending rural poverty in Africa. Natural Resource Perspectives, 101, 1-4, 2006.

[Cecchini03] Cecchini, S., & Scott, C. Can information and communications technology applications contribute to poverty reduction? Lessons from rural India. Information Technology for Development, 10(2), 73-84, 2003.

[Davies00] Davies, H. T., Nutley, S. M., & Smith, P. C. (2000). What works? Evidence-based policy and practice in public services. MIT Press, 2000.

[Deloitte14] Deloitte. Value of Connectivity. The Economic and social benefits of expanding internet access, February 2014

[Deloitte15] Deloitte. Broadband infrastructure sharing policies and strategies in emerging markets. APC, April 2015. Available at https://www.apc.org/en/node/20382

[EIU14] EIU (Economist Intelligence Unit). Global Microscope 2014: The enabling environment for financial inclusion. Sponsored by MIF/IDB, CAF, ACCION and Citi. EIU, New York, NY, 2014.

[Fedesarrollo13] Fedesarrollo. La calidad de la telefonía móvil en Colombia. Bogotá: Fedesarrollo, 2013.

[Flores10] Flores, E. & Mariscal, J. Oportunidades y desafíos de la banda ancha móvil en América Latina. Acelerando la revolución digital: banda ancha para América Latina y el Caribe. CEPAL, 2010.

[Galperin05] Galperin, H. Wireless networks and rural development: opportunities for Latin America. Information Technologies and International Development, 2(3), 47-56, 2005.

[Garcia10] García López, R., & Moreno, M. G. La gestión para resultados en el desarrollo. Avances y desafíos en América Latina y el Caribe. Washington, DC: Banco Interamericano de Desarrollo, 2010.

Page 7: TUCAN3G business model: case study and field verification

ICT-601102 STP Document number: D34 Title of deliverable: TUCAN3G business model: case study and field verification

TUCAN3G_D34EHASd.docx 7

[GSMA12] GSMA Development Fund and Cherie Blair Foudation for Women. Women & Mobile: A Global Opportunity A study on the mobile phone gender gap in low and middle-income countries. July 2012. Available online: http://www.cherieblairfoundation.org/women-and-mobile-a-global-opportunity/

[GSMA13] GSMA MECS & MIF. Beyond Coverage: The opportunity for mobile operators to improve access to energy in Latin America. GSMA’s Mobile Enabled Community Services and Multilateral Investment Fund, 2013.

[GSMA14b] Arese Lucini, B &Hatt, T. Country overview: Peru. GSMA Intelligence Analysis Report, February 2014. Available online: https://gsmaintelligence.com/research/2014/02/country-overview-peru/419/

[GSMA14] GSMA Public Policy Committee. Mobile Policy Handbook: An Insider’s Guide. GSMA, 2014.

[GSMA15] GSMA Intelligence. The Mobile Economy Global Report 2015. GSMA, 2015.

[Hatt15] Hatt, T., Okeleke, K., & Melon. M. Closing the coverage gap — a view from Asia. GSMA Intelligence, 2015.

[Intven00]

Intven, H. Telecommunications regulation handbook. World Bank, 2000.

[ITU14] International Telecommunication Union. World Telecommunication Development Conference (WTDC-14): Final Report. ITU, 2014.

[Jordan13] Jordán, V., &Peres, W. Banda ancha en América Latina: más allá de la conectividad. CEPAL, 2013.

[Kassam14] Kassam,A. Green Power for Mobile Bi-Annual Report 2014. GSMA, 2014.

[Katz12] R. Katz. The Impact of Broadband on the Economy: Research to Date and Policy Issues. ITU, April 2012

[Kelly11] Kelly, T., &Rossotto, C. M. (Eds.). Broadband strategies handbook. World Bank Publications, 2011.

[Kim10] Kim, Y., Kelly, T., & Raja, S. Building broadband: Strategies and policies for the developing world. World Bank Publications, 2010.

[Meddour11] Meddour, Djamal-Eddine, Tinku Rasheed, and YvonGourhant. "On the role of infrastructure sharing for mobile network operators in emerging markets." Computer Networks 55.7 (2011): 1576-1591.

Page 8: TUCAN3G business model: case study and field verification

8

[Medina14] F. Medina Telecomunicaciones y competencias: Políticas sectoriales exitosas y desafíos. Horizontal. Available at http://www.horizontalchile.cl/publicaciones/telecomunicaciones-y-competencias-politicas-sectoriales-exitosas-y-desafios/

[OECD14] OECD, “New Approaches to Spectrum Management”, OECD Digital Economy Papers, No. 235, OECD Publishing, 2014. http://dx.doi.org/10.1787/5jz44fnq066c-en

[Okeleke15] K. Okeleke, B. A. Lucini, T. Hatt, “Agri VAS: Market Opportunity and Emerging Business Models”, GSMA Intelligence, February 2015

[Osterwalder10]

A. Osterwalder, Y. Pigneur. Business Model Generation: A Handbook for Visionaries, Game Changers, and Challengers. John Wiley & Sons, Inc, 2010

[Palmer15] T. Palmer, R. Zelezny-Green, “Airtel Green SIM”, GSMA - Mobile for Development, March 2015

[Patton12] Patton, C., Sawicki, D., & Clark, J. Basic Methods of Policy Analysis and Planning 3rd Edition. Pearson, 2012.

[Savitz12] Savitz, A. The triple bottom line: How today's best-run companies are achieving economic, social and environmental success--and how you can too. John Wiley & Sons, 2012.

[Serra07] Serra, A., Figueroa, V., & Saz, Á. Modelo abierto de gestión para resultados en el sector público. Banco Interamericano de Desarrollo (BID). Centro Latinoamericano de Administración para el Desarrollo (CLAD), 2007.

[Shabalala07] Shabalala, D. B. Towards a Digital Agenda for Developing Countries. Research Paper, 13, 2007.

[Simanis12] Simanis, E. Reality check at the bottom of the pyramid. Harvard Business Review, 90(6), 120-125, 2012.

[Sprague14] Sprague, Kara, et al. "Offline and falling behind: Barriers to internet adoption." McKinsey & Company, Tech. Rep (2014).

[TUCAN3G-D23] ICT-601102 STP TUCAN3G, “Parameters and basic conditions for the market research and the business model”, deliverable D23, June 2013, available online: www.ict-tucan3g.eu.

[TUCAN3G-D31] ICT-601102 STP TUCAN3G, “Markey Survey”, deliverable D31, April 2014, available online: www.ict-tucan3g.eu.

[TUCAN3G-D41] ICT-601102 STP TUCAN3G, “UMTS/HSPA network dimensioning”, deliverable D41, November 2013, available online: www.ict-tucan3g.eu.

Page 9: TUCAN3G business model: case study and field verification

ICT-601102 STP Document number: D34 Title of deliverable: TUCAN3G business model: case study and field verification

TUCAN3G_D34EHASd.docx 9

[TUCAN3G-D51] ICT-601102 STP TUCAN3G, “Technical requirements and evaluation of WiLD, WIMAX and VSAT for backhauling rural femtocells networks”, deliverable D51, October 2013, available online: www.ict-tucan3g.eu.

[UN-DESA15] United Nations. Department of Economic and Social Affairs. Transforming our World: The 2030 Agenda for Sustainable Development. United Nations Publications, 2015.

[Williams12] C. Williams, G. Solomon, R. Pepper. “What is the impact of mobile telephony on economic growth? A report for the GSMA”, Deloitte-GSMA, March 2012

Page 10: TUCAN3G business model: case study and field verification

10

List of abbreviations & symbols 3G 3rd Generation (of mobile telecommunications technology)

ARPU Average Revenue Per User

BMC Business Model Canvas

BOP Bottom of Pyramid

CAPEX Capital Expenditure

EBITDA Earnings Before Interest, Taxes, Depreciation, and Amortization

GDP Gross Domestic Product

ICT Information and telecommunication technologies

HNB Home Node B (femtocell)

M2M Machine to Machine

M4D Mobile for Development

MNO Mobile Network Operator

NPV Net Present Value

NGO Non-governmental organization

OIMR Operador de InfraestructuraMóvil Rural

OPEX Operational expenditures

QoS Quality of Service

RNO Rural Network Operator

ROI Return On Investment

SDG Sustainable Development Goals

SMS Short Message Service

SROI Social Return On Investment

TCO Total Cost of Ownership

UN United Nations

VSAT Very Small Aperture Terminal

WiLD WiFi Long Distance

WiMAX Worldwide Interoperability for Microwave Access

Page 11: TUCAN3G business model: case study and field verification

ICT-601102 STP Document number: D34 Title of deliverable: TUCAN3G business model: case study and field verification

TUCAN3G_D34EHASd.docx 11

1 INTRODUCTION

Building sustainable telecommunications infrastructure to provide 3G mobile services for underserved and isolated rural communities in developing countries

1.1 Problem statement

Living conditions in isolated rural communities are hard because they lack access to the basic services, i.e.: electricity, water, sanitation, health and education. That is the case in Peruvian Amazon, where many rural communities are only connected to the nearest city by riverboat, a journey that can take up to 3 days. People living in such remote and rural areas of developing countries are at the Bottom-of-Pyramid (BOP); they are neglected as a target market, even by public services. However, the United Nations’ Sustainable Development Goals (SDGs) [UN-DESA15] recognize the benefits of access to information and telecommunication technologies (ICTs) for spurring economic growth and social inclusion. ICTs are key enablers of sustainable development, yet the rural-urban digital divide needs to be solved. The reason for the rural-urban digital divide is three-fold. First, difficult rural geography hampers the deployment of network infrastructures that enable basic service delivery. Furthermore, rural populations in developing countries have high poverty incidence rates making them an unprofitable customer segment. The combination of high infrastructure deployment costs and low expected revenues leads to no profits for business. In the case of rural telecommunication network infrastructures, the wireless or mobile technologies could be an appropriate solution when fixed broadband networks cannot be deployed. However, the mobile operators still do not find attractive business opportunities in low-density areas where rural communities have less than 1.000 inhabitants. Consequently, utility and telecom companies are reluctant to provide services for the under-served poor, unless they are required or subsidized by the Government. Despite mobile operators are concerned about user demand in BOP markets, an analysis of the mobile services usage trends in rural areas of developing countries shows that these are attractive market opportunities. As an example, we estimate the market size in the rural areas of Peru and Colombia.

1.2 Mobile service demand in rural areas

The total addressable rural mobile broadband market is an estimation of the number of people who live outside of the range of 3G networks and live in rural communities. The value of the addressable rural mobile broadband market was then estimated by multiplying the addressable market population by region specific annual average return per user (ARPUs). The mobile telecommunications industry is one of the largest and fastest growing industries in the world. According to GMSA Intelligence [GSMA15], mobile subscriptions are growing at a compounded annual growth rate (CAGR) of 4% globally. Meanwhile, mobile-data traffic is expected to grow at a CAGR of 46% between 2014 and 2019. Furthermore, GSMA estimates the unique subscriber penetration globally to be only 50% percent. The Table 1 below shows the evolution of ICT access and use indicators between 2010 and 2015 in Peru and Colombia, according to the ITU-T and compared to other developing countries.

Page 12: TUCAN3G business model: case study and field verification

12

ICT access and use indicators

2010 2011 2012 2013 2014 2015*

Fixed-telephone subscriptions per 100 inhabitants Peru 10,0 10,2 10,5 10,5 9,8 9,6 Colombia 15,5 15,1 14,8 14,8 14,7 N/A Developing 11,9 11,5 11,2 10,6 10,0 9,4 World 17,8 17,2 16,7 15,9 15,2 14,5 Mobile-cellular telephone subscriptions per 100 inhabitants Peru 98,3 108,3 97,0 97,2 102,9 108,0 Colombia 95,8 98,1 102,9 104,1 113,1 N/A Developing 68,5 77,4 82,1 87,8 91,1 91,8 World 76,6 83,8 88,1 93,1 96,1 96,8 Percentage of households with a computer Peru 23,4 25,4 29,9 32,0 32,3 N/A Colombia 26,1 29,9 38,4 42,2 44,5 N/A Developing 22,6 25,1 27,3 29,2 31,0 32,9 World 35,8 37,8 40,0 41,8 43,6 45,4 Percentage of households with Internet access Peru 13,0 16,4 20,2 22,1 23,5 N/A Colombia 19,3 23,4 32,1 35,7 38,0 N/A Developing 16,4 20,5 24,2 28,6 31,5 34,1 World 29,9 33,6 37,1 41,2 43,9 46,4 Percentage of individuals using the Internet Peru 34,8 36,0 38,2 39,2 40,2 N/A Colombia 36,5 40,4 49,0 51,7 52,6 N/A Developing 16,4 20,5 24,2 28,6 31,5 34,1 World 29,9 33,6 37,1 41,2 43,9 46,4 Fixed-broadband subscriptions per 100 inhabitants Peru 3,1 4,0 4,7 5,3 5,7 6,3 Colombia 5,7 7,1 8,3 9,4 10,3 N/A Developing 4,2 4,9 5,4 6,2 6,6 7,1 World 7,6 8,4 9,0 9,9 10,3 10,8 Active mobile-broadband subscriptions per 100 inhabitants Peru 0,9 1,3 2,5 2,8 44,6 49,7 Colombia 5,6 3,7 5,1 25,1 45,1 N/A Developing 4,5 8,3 12,4 17,4 27,9 39,1 World 11,5 16,7 21,7 27,3 37,2 47,2

Table 1: ICT Development Index (IDI) indicators (source ITU-T) While mobile adoption and use continues to grow, MNOs are racing to expand network coverage to the last remaining untouched markets. GSMA and McKinsey [Sprague14] estimates that 64% of people without access to mobile coverage, or about 900 million people, live in rural communities. With a global annual ARPU of $146 (annual), the addressable market opportunity for rural mobile services is over $130 billion. The largest regional market opportunity is in Asia ($61 billion). Meanwhile, Latin America boasts an estimated $14 billion market opportunity. Furthermore, according to ITU’s ICT Facts and Figures 20151, more than 2.4 billion people living in rural communities worldwide do not have access to 3G services. The total market value of 3G mobile services in underserved rural areas would be over $350 billion.

1 https://www.itu.int/en/ITU-D/Statistics/Pages/facts/default.aspx

Page 13: TUCAN3G business model: case study and field verification

ICT-601102 STP Document number: D34 Title of deliverable: TUCAN3G business model: case study and field verification

TUCAN3G_D34EHASd.docx 13

Total population

Underserved rural population

Annual ARPU

Market value (in millions)

Africa 1111 227 $ 96 $ 21,771 Asia & Pacific 4427 565 $ 108 $ 61,046 Latam 600 108 $ 132 $ 14,193 World 7125 903 $ 146 $ 131,640

Table 2: Mobile market size and value estimations The deliverable [TUCAN3G-D31] presented evidence concerning the demand of mobile services in the target areas. It was found that 68.9% of Peruvian households surveyed have one member with a basic phone, despite their relatively low monthly income (less that USD 140 for 42% of the population, which in many cases is not a fixed monthly income, although it is above the national poverty line). According to the survey, what most discourages the customer base from using mobile network services is the poor quality of their infrastructure. Actually, in rural areas, the network tends to drop regularly with power cuts or lack of electricity to be frequent. About 72% of the population has access to electricity in their home, but only in a restricted basis (few hours a day) while the distances are long to travel in order to get access of the mobile network, especially when transportation costs are high. Similar facts are described in [TUCAN3G-D31] for the rural areas in Colombia, which are not a target in the TUCAN3G pilot but they will be considered for later scaling up of the solution. Most of the Colombian households surveyed earn a similar amount of a varied monthly income and have one basic phone. However, contrary to the Peruvians, the majority of the Colombians recognize the advantages of using internet more often than once per month. The main reasons for the low use of internet are lack of knowledge about the use of technology, the small number of access points, which are mostly far away from their residence, the instability of the network due to electricity failures and high costs. Peru mobile market estimate According to the ICT Development Index (Figure 1), Peru has grown much more in the use of mobile broadband than in fixed-broadband, following the same trend than in other developing countries.

Page 14: TUCAN3G business model: case study and field verification

14

Figure 1: ICT Development Index (Peru 2010-2015) Data provided by the national statistics office -INEI (Instituto Nacional de Estadística e Informática)- (Table 3) shows that rural areas in Peru are quite behind the national average in both access and use of internet. Surprisingly, an estimated 11.5% of total people living in rural areas is using internet even though they do not have access at home. They usually travel to the nearest city to find a cybercafé or public library where they can use internet access. The percentage of households with a computer is also quite low and therefore, the mobile phone would be the best device to provide internet access in isolated rural communities.

Internet access and use indicators

2010 2011 2012 2013 2014

Percentage of households with Internet access Rural 0,3 0,4 0,8 0,9 1,2 Peru 13,0 16,4 20,2 22,1 23,5 Percentage of individuals using the Internet Rural 9,9 10,0 10,4 10,9 11,5 Peru 34,8 36,0 38,2 39,2 40,2 Percentage of households with a computer Rural 2,6 3,5 4,4 5,8 6,1 Peru 23,4 25,4 29,9 32,0 32,3

Table 3: Internet access and use indicators (Source: INEI) The INEI also estimates (Table 4) that there are more than 7 million inhabitants living in rural areas; bridge the urban-rural gap (more than 30%) by providing 3G mobile services means a market opportunity, at least, of 2 million potentially new customers (mobile internet new subscriptions). With a monthly ARPU of $9 (Source: [GSMA14b]), this is a market opportunity for mobile operators estimated at least $216 million annual.

Page 15: TUCAN3G business model: case study and field verification

ICT-601102 STP Document number: D34 Title of deliverable: TUCAN3G business model: case study and field verification

TUCAN3G_D34EHASd.docx 15

Rural population 2005 2010 2012 2013 2015 TOTAL 8.028.132 7.656.096 7.500.133 7.420.750 7.257.989 Men 4.157.367 3.984.556 3.911.671 3.874.473 3.797.808 Women 3.870.765 3.671.540 3.588.462 3.546.277 3.460.181

Table 4: Peru rural population estimates by gender, 2005-2015 (Source INEI) In Peru, 30.000 localities lack mobile service according to FONIE-FITEL2. The vast majority of these localities (85.5%) have fewer than 100 inhabitants. In 2016, the FONIE is involved in 663 districts3 located nationwide benefiting more than 5 million inhabitants that live in the poorest areas of Peru. The solution proposed by TUCAN3G to provide 3G mobile telephone and internet services could then have a positive social impact and improve the life conditions for at least 4 million Peruvian people living in poor rural communities with population under 100 inhabitants. Colombia mobile market estimate In Colombia, internet adoption is growing. In the latest 2015 quarter, there were 5,258,113 broadband connections and a total 4,854,509 (3G and 4G) mobile broadband subscriptions.

Figure 2: Broadband connections in Colombia (Source: MinTIC 2015)

Since 2012, the total number of mobile internet subscriptions has been growing at more than 20% annual, up to 27 million users at the end of 2014. That is more than half of mobile telephone subscriptions; about 80% of internet mobile subscriptions are prepaid, but there are more than 5 million post-paid service contracts. The growing rate of mobile post-paid subscriptions in 2012-2014 was about 37% annual.

2012 2013 2014 Fixed internet 3.906.885 4.497.678 5.051.552 Mobile internet - prepaid 15.687.971 14.676.422 21.412.556 Mobile internet - post-paid 3.209.059 4.563.644 5.565.663 Fixed telephone 7.030.348 7.133.260 7.180.937 Mobile telephone 49.066.359 50.295.114 55.330.727

Table 5: Internet and telephone subscriptions in Colombia, 2012-2014

The main mobile operators in Colombia are Claro, Movistar, and Tigo. Claro reported that by December 2014 their mobile coverage reached 1.100 municipalities out of 1.123 in Colombia, whereas

2http://www.midis.gob.pe/dmdocuments/fonie_taller_04model_telefonia_movil_fitel.pdf 3http://www.midis.gob.pe/dmdocuments/DISTRITOS_FOCALIZADOS.pdf

Page 16: TUCAN3G business model: case study and field verification

16

Movistar was in 961 municipalities and Tigo in 565. Andean and Caribbean are the best-connected regions in Colombia, whereas Orinoquía and Amazonía are the most unconnected regions in the country. These underserved regions have vast rural areas with access difficulties, due to geographic conditions. Pacific region comprises the following departments: Chocó, Valle del Cauca, Cauca y Nariño. In this region settled more than eight million people, 4.613.377 are located in Valle del Cauca, 500.076 in Chocó, 1.379.070 in Cauca, and 1.744.275 in Nariño. Valle del Cauca is one of the most developed departments in the country. In Choco, 70% of the department are rural areas, 49% of its population is located in main zones, and 51% in rural zones. In Nariño, 49% of its population is located in main zones and 51% in rural zones. Finally, in Cauca, 40% of its population lives in urban zones and 60% in rural zones. Orinoquia region has near to 1.662.894 inhabitants in four departments: Arauca, Casanare, Meta and Vichada. Arauca department has an estimated population of 265.160 inhabitants for 2016 (Dane 2016). A 63% of this population lives in the main city and 37% in rural zone. Casanare department has an estimated population of 362.698 inhabitants for 2016, a 68% live in urban zones and 32% in rural zones. Meta department has a population of 961.334, a 64.6% live in urban zones and 35.4% in rural zones. Vichada has a population of 73.702 inhabitants, 57% rural and 43% urban, divided into four municipalities: Puerto Carreño (main city), La Primavera, Santa Rosalía y Cumaribo. In addition to these municipalities, Vichada department has several “resguardos” (indigenous settlements), which have much relevance in the region. On the contrary, to Pacific region, Orinoquía has a majority of population in urban zones or main cities, where mobile coverage conditions are better, and less percentage of people in rural zones. However, it is important to take into account that these departments have big areas, therefore the population spreading and the difficult conditions to move towards an urban zone are reasons for improving the coverage conditions in the whole territory. Amazonia Region has a population of 1.024.915 inhabitants in six departments: Putumayo, Caquetá, Guaviare, Guainía, Vaupés and Amazonas. In Putumayo department, there are 349.537 inhabitants; a 48% live in urban zones and 52% in rural zones. In Caquetá live a total population of 447.642 inhabitants; a 59% were located in urban zones and 41% in rural zones. In Guaviare department, there are 110.011 inhabitants; 59% are located in urban areas and 41% in rural zones. The fixed telephone service has very low levels of adoption. In Guainía, there are 41.482 inhabitants; 31% live in urban zones and 69% in rural zones, while in Vaupes department live 43.665 inhabitants; 39% live in rural zones and 61% in rural zones. In Mitu, Vaupes main city, fixed telephone service is offered by Telecom, and Claro, Movistar and Tigo offer mobile phone service. Amazonas department get a population near to 76.243 inhabitants, 37% live in urban zones, and 63% live in rural zones.

Region Total population Rural population Pacific 4,613,377 3,997,334 Orinoquia 1,662,894 596,495 Amazonia 1,024,915 420,397

Total 5,014,226

Table 6: Rural population in Colombia According to the Encuesta Nacional de Calidad de Vida (ECV 2015), the percentage of individuals in Colombian rural areas using internet is a 31.2% and 62.9% in urban areas, so the urban-rural gap is about 30%. As in the case of Peru, the percentage of households with a computer is quite low; therefore the mobile phone would be also the best device to provide internet access in isolated rural communities (a 79.1% of individuals in Colombian rural areas are using mobile telephony). With a monthly ARPU of $9, bridge the urban-rural gap by providing 3G mobile internet services is a market opportunity for mobile operators estimated at least $162 million annual.

Page 17: TUCAN3G business model: case study and field verification

ICT-601102 STP Document number: D34 Title of deliverable: TUCAN3G business model: case study and field verification

TUCAN3G_D34EHASd.docx 17

1.3 Appropriate technology solution

Since there is an attractive potential market, the key factor to drive business is to reduce the cost of building access and transport networks in remote rural areas. We found that population centres in the Peruvian Amazon jungle are not attractive to mobile operators, because of low-income population, inaccessibility via wired terrestrial networks, lack of reliable electricity supply and harsh weather conditions to ensure quality of service (QoS). The current solution to solve the mobile coverage gap in isolated rural areas in the Peruvian Amazon is to deploy satellite-based rural telephony and internet, but this is very expensive for the majority of users. Satellite links have been the preferred technology for most MNOs as a rural backhaul solution due to the relatively low installation and equipment costs (CAPEX) and short installation time. However, satellite backhaul carries a heavy operating expense (OPEX) which could hinder the profitability and sustainability of the infrastructure in low-income and low-density rural communities.

The TUCAN3G project has developed an innovative technical solution for affordable 3G mobile network infrastructures in rural areas of developing countries. Part of the value proposition of TUCAN3G business model for a MNO is the cost savings when compared to the satellite solution. By so doing, its social value proposition is more likely to be achieved. TUCAN3G will enable to provide mobile services profitably in rural population centres with a few hundreds of users. The network architecture of the TUCAN3G solution is based on a terrestrial wireless backhaul network and small cells as mobile access nodes, comprising the following elements: • Antenna Support Structure comprises of radio tower and other mounting elements designed to

support antennas and telecommunications equipment. • Solar PV System is an autonomous photovoltaic system designed to supply electric power for

operating both femtocells and backhaul communications equipment. • Wireless IP Subsystem providing IP connectivity (radio links) between nodes of the backhaul

network and the interface link between the backhaul gateway and the core network. This telecommunication equipment is installed in the antenna towers.

• Radio Network Subsystem (RNS) consists of a radio network controller (RNC) and a set of

femtocells providing the (3G/4G) connection between mobile phones and telecommunication service providers.

• Core Network Support Systems manage configuration, supervision and performance of all

network elements. The RNC in the radio network subsystem establishes a standard interface link with elements of the operating and business support systems (OSS/BSS) in the core network.

A femtocell (also known as Home Node B - HNB) performs many of the functions of a Node B, but it is optimised for deployment in the indoor premises and small coverage public hotspots. The femtocell concept was originally conceived for residential environments/people homes. However, it has evolved to include other usages such as enterprise and public hotspots. The solution proposed in TUCAN3G is based on outdoor femtocells (HNB) base stations for the mobile access network in isolated rural areas. The big advantage of femtocells is that they are much more cost-effective to provide mobile services when there are fewer simultaneous users. The density of mobile traffic is the driving force behind macro cell advantage. A small cell based solution is particularly advantageous in low population density areas, as is very much the case in isolated rural areas. Femtocells-based deployment of mobile

Page 18: TUCAN3G business model: case study and field verification

18

network services requires access to licensed spectrum. Therefore, the role of the MNO is vital in order to deliver mobile services in rural isolated communities.

The other alternative option reducing the total cost of ownership is a multi-hop terrestrial wireless backhaul network to connect the remote site to the core mobile network. The deployment of terrestrial wireless networks has a relatively low OPEX. However, there needs to be a significant initial investment to acquire and build the network infrastructure. The CAPEX depends mostly on the size of the steel tower required to house the ICT equipment. The towers range between 15m and 120m depending on the network site’s geography because the radio links need to be in direct line-of-sight.

1.4 Social impact of mobile services

Finally, the TUCAN3G project aims at creating social value dramatically changing the lives of people living in isolated rural communities. The use of ICTs provides an opportunity to tackle many of the challenges in the international development agenda. According to the World Bank, a 10% increase in mobile and broadband penetration increases per capita GDP by 0.81% and 1.38% respectively in developing countries [Bhavnani08] [Katz12] [Williams12]. Access to mobile networks enables people to access information and services that were previously non-existent. Unfortunately, the supply of mobile services does not satisfy the demand in remote and rural areas where the coverage gap is huge. The TUCAN3G project is aligned with The Global Goals4 for Sustainable Development as defined in the [UN-DESA15]. In partnership with the public, social and private sectors, we aim at expanding access to and use of mobile broadband services, particularly to people living in isolated rural areas of developing countries. People should have the same opportunities in rural communities as those living in urban areas. Actions must be taken for the elimination of inequality, promote rural development and make a positive change in people’s lives. This implies to promote inclusive economies in which underserved poor people have access to basic infrastructures, such as telecommunication networks facilitating education, health and financial services. Below, we outline the typical mobile applications that are most relevant to the needs of the rural poor according to GSMA Mobile for development (M4D) initiative5.

4 www.globalgoals.org 5 www.gsma.com/mobilefordevelopment/

Figure 3: Architecture for the Tucan3G demonstration platform

Page 19: TUCAN3G business model: case study and field verification

ICT-601102 STP Document number: D34 Title of deliverable: TUCAN3G business model: case study and field verification

TUCAN3G_D34EHASd.docx 19

Mobile for Basic Infrastructure Services The majority of the areas that lack access to telecom infrastructures also lack access to grid electricity, clean water, and sanitation systems. Deploying an ICT infrastructure has the potential to improve residents’ access to other utilities through infrastructure sharing or M2M applications. In Peru, an estimated 2.9 million people have access to mobile networks, but lack electricity. One potential solution to address the energy-starved population would be to use the telecom infrastructure to provide electricity as well as mobile coverage. Most of the towers in rural and isolated areas are off-grid and therefore must run off solar energy. By installing extra solar panels to each tower site, the tower operator could generate enough energy for the tower as well as for the local residents to charge their electronic devices. The operator could distribute the power via a mini-grid, transferable batteries, or direct charging at the tower site. Mobile for Women Empowerment Women represent nearly two thirds of the untapped growth in the global mobile market and are 21% less likely to own a mobile phone than a man is. According to GSMA Intelligence, only 30% of women in rural areas of Peru have access to mobile phones. This leaves more than 2 million Peruvian women in rural areas that lack access to mobile services. Studies published by GSMA and Cherie Blair Foundation [GSMA12] have shown mobile ownership has a clear effect on female empowerment. According to the GSMA, 90% of women feel safer because they have a mobile phone and 85% of women reported feeling more independent because of their mobile. Mobile for Agriculture Agriculture accounts for 11% of the GDP in developing countries. According to the latest census material, farmers account for more than half of the labour force in both the Napo and Balsapuerto districts in our pilots. mAgri applications have the ability to mitigate the information gap faced by rural farmers in developing countries. According to a 2014 Deloitte study, access to the internet can increase a farmer’s profit by nearly 33%. One of the most successful mAgri applications is Airtel’s Green SIM in India. Over 3 million Indian farmers have subscribed to the Green SIM mobile service. Green SIM sends farmers voice-based messages on best farming practices, weather patterns, and market prices. In addition to daily voice messages, Green SIM offers a farmer helpline, connecting farmers to experts who can offer advice on plant protection, fertilizer application and much more. Mobile Money According to the Global Microscope 2014 report, published by the Inter-American Development Bank, Peru ranked the most conducive market in the developing world for expanding access to financial services. In 2013, Peru passed Law No. 29985, which legalized the transfer of money by non-banking institutions on electronic platforms. The law was created with the hopes of encouraging financial inclusion. Although the regulatory environment is excellent for combating financial exclusion, 80% of Peruvians above the age 15 still do not have a bank account. Of the 17 million unbanked or underbanked adults in Peru, 70% of them have access to a mobile phone. The most notable mMoney application in the world, which has not reached Peru, is M-Pesa. M-Pesa allows users to send and receive money to friends, families, and businesses. M-Pesa and companies like it have enabled family members who left their rural homes to send remittances back to their family. Furthermore, mMoney has increased the efficiency of microfinance institutions operating in rural areas where clients are spread over hundreds of kilometres.

Page 20: TUCAN3G business model: case study and field verification

20

2 METHODS AND ASSUMPTIONS This chapter describes the assumptions and methods behind the business case analysis and business model validation. We define the financial metrics that will be compared in different scenarios to support decision-making. Certain assumptions will be made to simplify and clarify the analysis of results.

2.1 Financial metrics

The main objective of our business case analysis is to help MNOs, potential investors and governments to decide if the investment in rural network infrastructure is financially sustainable. More specifically, we would like to compare two scenarios: (a) the current solution based on satellite and (b) the TUCAN3G solution approach based on a terrestrial wireless network. A simple way to compare is using Total Cost of Ownership (TCO), which is the sum of Capital Expenditures (CAPEX) and Operating Expenses (OPEX) in a period of 5 years. In the initial capital investment (CAPEX), we include both the cost of acquisition for the network elements and the cost of transportation and installation. In the operating costs (OPEX), it is included the network operation and maintenance.

𝑇𝑇𝑇𝑇𝑇𝑇𝑛𝑛 = 𝑇𝑇𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 + �𝑇𝑇𝐶𝐶𝐶𝐶𝐶𝐶𝑡𝑡

𝑛𝑛

𝑡𝑡=0

The cost savings, calculated as the difference of TCO between both scenarios, can be used to decide the best solution since the revenue estimate (mobile service demand) does not depend on the backhaul technology used. However, we need to estimate revenue to verify the financial sustainability. The Revenue forecast depends on two factors: Average Revenue Per User (ARPU) and Mobile Phone User Penetration. The mobile phone user penetration is the percentage of total inhabitants (population) that have a mobile service subscription. Estimated sales revenues are simply the result of multiplying the ARPU by the number of users. The number of users is the determined by the total number of inhabitants (population) and the mobile phone user penetration rate. From the estimated Revenue and OPEX in a period, usually between 3 and 5 years, we can calculate a Gross Profit estimate. Then, we can calculate the Return On Investment (ROI) as the Gross Profit (return) divided by the CAPEX (cost of investment) expressed as a percentage or ratio. The ROI metric is very easy to calculate and to interpret; must be positive and higher than other investment alternatives. However, one limitation of using ROI is that do not consider the Time Value of Money. The Net Present Value (NPV) is the financial metric commonly used to analyse the profitability of an investment or project. A positive NPV would indicate that the investment creates value for an amount equal to the estimated net present value. The formula to calculate the NPV of a stream of cash flows (Ci) as of n periods from now generated by an investment (I) with a discount rate (r) is:

𝑁𝑁𝐶𝐶𝑁𝑁 = �𝑇𝑇𝑖𝑖

(1 + 𝑟𝑟)𝑖𝑖

𝑛𝑛

𝑖𝑖=1

− 𝐼𝐼

We will use a value of r = 9%, which is the Social Discount Rate applied in public infrastructures projects in Peru, and is related with the interest rate required for this kind of investments. Finally, a Discounted Payback Period gives the number of years it takes to break even from the initial investment, considering the time value of money. In contrast to the (non-discounted) payback period, which only measure how long it take for the initial cash outflow to be paid back, we use discounted cash flows and calculate when the NPV becomes positive (break-even).

Page 21: TUCAN3G business model: case study and field verification

ICT-601102 STP Document number: D34 Title of deliverable: TUCAN3G business model: case study and field verification

TUCAN3G_D34EHASd.docx 21

2.2 Assumptions

Radio tower height and steel price

The supporting infrastructure for antennas and communications equipment is usually a tower, made of a steel lattice structure attached to the ground that provides strength, low weight and wind resistance. A tower supports antennas at a height where they can transmit and receive radio signals; the use of WiFi and WiMAX in multi-hop backhaul networks is feasible only and only if LOS (line-of-sight) is guaranteed between the two nodes of every link. Every location where a backhaul node is installed needs a mast or tower on which antennas and communications systems will be installed. For most scenarios (mainly in jungle), the cost of towers is very dominant over all the other cost components, because the LOS requirement implies the use of large supporting structures to make possible the placement of antennas at the right elevation. The cost of a tower is roughly proportional to the square of its height; normal values are 15, 30, 45, 70 and 90 meters. Highland site locations usually have 15 meters, whereas jungle site locations may require 90 meters towers. A guyed mast is cheaper to build than a self-supporting tower of equal height and needs additional land to accommodate the guy-lines, and is thus best suited to rural locations where land is relatively cheap. A steel lattice tower is cheaper to build than a concrete tower of equal height. However, it is important to notice that the highest towers are more sensible to the fluctuations of prices in the steel markets. The London Metals Exchange (LME) is the world’s premier metal exchange. Prices derived from trading on the Exchange are used as reference prices in the majority of the world’s physical metal markets.

Figure 4: Historical data of LME Steel Billet prices

Our analysis of financial results will assume the worst (most expensive) case, where radio towers are built in a jungle area with an average height of 75m. The prices of steel and cost of the equipment comprising the antenna support structures (radio tower) that we use in our analysis is provided by the GTR-PUCP. It is important to notice that these data are an estimate based on previous rural network deployments along the last 20 years. The radio tower used during the TUCAN3G pilot were already built, they are now being use for installing new 3G mobile services. Multi-hop networks and costs per site

In a multi-hop wireless network, there may be intermediate nodes acting as repeaters, but not providing any end-user service. To estimate the cost per node, the intermediate node costs must be

Page 22: TUCAN3G business model: case study and field verification

22

attributable to one of the two adjacent nodes. However, in our analysis, we assume there are no intermediate nodes, since most rural communities are reachable in one hop using WiLD (or VSAT). Regardless of the technology used for the backhaul network, in any scenario there are direct costs associated to each site location (network node) and other indirect fixed costs attributable to the network head end and gateway to the core network. In estimating total costs per node, those indirect costs are distributed among all nodes in the network. Therefore, the more network coverage, the lower the attributable costs. In the TUCAN3G pilot, the network coverage comprises six rural communities. However, our analysis will assume that the network has 400 nodes or site locations so that the fixed indirect costs (Radio Network Controller installed at the headend) are weighted as in a real case. Revenue forecast

An easy way to estimate mobile revenue is by multiplying the expected number of mobile users by the ARPU. According to [GSMA14b], the ARPU in Peru is estimated about $ 9 monthly with a compounded annual growth (CAGR) at 5%. GSMA also estimates a subscriber growth of 5%, giving a total revenue growth of 10%. The estimation of backhaul bandwidth requirements in deliverable D41 forecasted a much more rapid grow at CAGR 16% during the first 5 years. It is reasonable that for network planning the forecast is much more optimistic, to ensure the quality of service. However, in our financial analysis we prefer to be more realistic. During the first years, the demand for mobile services will grow steadily as usually happens in urban areas and in rural areas of other developing countries. As we said in section 1.2, data provided by ITU-D indicates that in developing countries there is an average about 90 mobile telephony users and 40 mobile internet users per 100 inhabitants. However, in very isolated rural areas, we have observed that the population is declining (people leave their communities for a better living and job opportunities). Therefore, we will assume a 5% revenue growth as a more conservative value in our calculations. We also conducted a field survey in several locations of the Peruvian Amazon jungle to have an initial estimate of ARPU in such isolated communities. The analysis of results from TUCAN3G pilot gives us an estimate average of 27 users per 100 inhabitants and $ 11 of monthly expenses per mobile user. Obviously, this estimate does not consider the revenues from incoming calls, and the credit available (not spent) in prepaid accounts, but we can use it as estimation of ARPU. During the TUCAN3G pilot, we have been monitoring the real traffic data to measure the minutes of voice calls and usage of data connections. From the traffic data provided by the network controller, we validated our estimate of annual revenues multiplying the voice traffic (minutes) and data traffic (MB) by an average revenue per minute/MB. For this validation, we adjusted the estimated number of users, since San Gabriel de Varadero is a big location with pre-existing 2G mobile users and we did not have access to this traffic data (2G and 3G were simultaneously offered during TUCAN3G pilot). Since the average population of the 6 pilot localities in TUCAN3G is 338 per location, we assume 91 mobile users per site and annual revenue about $12,000 per site. Finally, our calculations are based in real traffic measures, but the TUCAN3G pilot has been only working during 2 months and the collected data could change in coming months. Other financial parameters

Instead of building their own radio towers, it is quite common for mobile operator to rely on existing wireless infrastructure operators (TowerCo such as American Tower). When this infrastructure is shared, there is no cost of investment attributable to radio towers (CAPEX), but the cost of lease must be included (OPEX). Estimating the value of rural tower sharing expenses is difficult because the wireless rural backhaul market is nascent and lease rates of radio tower vary greatly. Tower lease rates are calculated based on a variety of inputs such as the geographic location, number of tenants, tower cash flow, the likelihood of future proximate towers and the negotiation skills of the tower owners. Lease rates can vary from a few hundred dollars per month to a few thousand dollars per month.

Page 23: TUCAN3G business model: case study and field verification

ICT-601102 STP Document number: D34 Title of deliverable: TUCAN3G business model: case study and field verification

TUCAN3G_D34EHASd.docx 23

The maximum limit is the marginal revenue per site expected by the mobile operator. On the other hand, a minimum acceptable for the tower owner would be such that covers the operating expenses. In our analysis, we assume a lease rate that allows the mobile operator a 40% of gross profit margin. This income allows the rural network provider to cover operating expenses and a payback period about 10 years with at least two housed tenants (as we will see later). Other assumptions used in the financial calculations are that the discount rate is 9% (as explained before) and taxes are 28% of the operating profit. In addition, other indirect costs such as Sales, General and Administrative costs (SGA) are estimated as a percentage of revenue (5%). Finally, we consider the tower asset's life is more than 25 years (4% annual depreciation), the batteries in the solar PV system last about 3 years and the solar panels have an expected life about 10 years. The femtocells in the jungle have a lower life expectancy, about 4-5 years (20-25% depreciation). All the other network equipment has depreciation at 10% (10 years life). Generally, the OPEX for network equipment other than radio towers is estimated at 10% of their cost of acquisition. The OPEX does not include the costs of transportation and installation; these are included in the CAPEX.

2.3 Scope

The financial model analysis refers to a “typical mobile site” located in an isolated rural community in the Amazon jungle,, although we also compared the total costs for the case of a rural community in the highlands. Our calculations are based on averaged revenue and costs obtained from the pilot. We have preferred to show results per location instead of summing revenue and costs of all the sites that mobile network covers, in order to analyse what implies to provide service in each new community. As previously mentioned, the averaged values of expenses were validated from real cases, previous rural network deployments performed by GTR-PUCP. As said above, the revenue and users profile were first estimated from field survey conducted to establish the baseline (ex-ante), where; the representative sampling size was calculated based on the estimated population of each location. The localities where the baseline survey was conducted are San Juan de Armanayacu, San Gabriel de Varadero, Huitotos de Negro Urco and Tutapishco. Typical users for the mobile services to be offered are people between 12 and 50 years old, both male and females. The population of these locations, according to the latest census available (Censo INEI6 2007) is as follows:

Table 7: Population in the population centres of TUCAN3G pilot (INEI 2007)

Department Province District Locality Pop.2007 Loreto Alto Amazonas Balsapuerto San Juan de Armanayacu 47 Loreto Alto Amazonas Balsapuerto San Gabriel de Varadero 780 Loreto Maynas Napo Huitotos de Negro Urco 263 Loreto Maynas Napo Tutapishco 287

TOTAL 1377

To calculate the sample size, we consider a simple design and the following considerations have been taken into account to determine the sampling size:

• The proportion of rural population between 10 and 54 years in Loreto region is 63%, with an estimated CAGR at -0,46%, according to data provided by INEI.

6Instituto Nacional de Estadística e Informática

Page 24: TUCAN3G business model: case study and field verification

24

Table 8: Rural population in the Loreto region

• The estimated population between 10 and 54 years in the five locations, according to the data provided by “Censo INEI 2007” and using the CAGR to calculate the projection in 2015 is:

Table 9: Estimated rural population in the five population centres Locality Pop.2007 Pop.2015

San Juan de Armanayacu 47 47 San Gabriel de Varadero 780 776 Huitotos de Negro Urco 263 262 Tutapishco 287 286 TOTAL 1377 1371

• The average size of households is 4 family members. • The sample size is estimated for each location, based on the above considerations, with a

confidence level of 95%, error margin 5% and selection probability 50%. The sample size was adjusted with a no answer factor at 2%.

Table 10: Determination of the sample size for the field survey District Locality Pop.2015 #Households #Sample

Balsapuerto San Juan de Armanayacu 47 12 6 Balsapuerto San Gabriel de Varadero 776 194 105 Napo Huitotos de Negro Urco 262 65 35 Napo Tutapishco 286 71 39 TOTAL 1371 342 185

Later, we validated our revenue estimate from real traffic data measures obtained during the TUCAN3G pilot. The TUCAN3G pilot provided mobile network services by deploying a 3G mobile access and a wireless backhaul platform in two distinct isolated rural areas in the Loreto region of Peruvian Amazon. At the Napo river area, specifically at the Maynas province (target area), there are four rural communities: Libertad, Negro Urco, San Luis de Tacsha Curaray and TutaPishco. The second target area located at the Paranapura river region, specifically at the Alto Amazonas province, includes two rural communities: San Juan and San Gabriel de Varadero. Traffic data was measured during more than two months, in our analysis we estimate revenue using the weekly average.

2.4 Scenarios

We based our definition of scenarios on the Business Model Canvas (BMC)[Osterwalder10]. The obvious business model design is when a MNO makes an investment to build its own (transport and access) network infrastructure. Extending mobile services to rural areas in developing countries is a major challenge and financial risk for the mobile operator. They often find it hard to justify network

Page 25: TUCAN3G business model: case study and field verification

ICT-601102 STP Document number: D34 Title of deliverable: TUCAN3G business model: case study and field verification

TUCAN3G_D34EHASd.docx 25

deployment cost when reaching remote communities with low population density. The expected revenues for the mobile operator are minimal and they usually do not justify the investment in equipment for a new location. Sometimes, operators deploy mobile network services infrastructure in difficult to reach rural areas due to government requirements, as condition for the authorization of their operator license. Their losses are leveraged with the revenue from other urban or more populated areas. Depending on the backhaul technology used, the first option of business model design leads us to two different scenarios in our financial analysis: satellite (current approach) and terrestrial wireless (as proposed in TUCAN3G network architecture).

Figure 5: Business model canvas for scenario in which MNO deploys new infrastructure A more attractive option for a business model design is sharing an existing backhaul network. Some local governments and NGOs have already deployed wireless networks to provide a communications infrastructure for their operations in rural and developing areas. For example, EHAS and GTR-PUCP have installed 144 radio towers in the Peruvian highlands and jungle to provide ICT services to municipalities, police stations, health posts, and schools.

Figure 6: Business model canvas for scenario in which backhaul provider leases transport network

Page 26: TUCAN3G business model: case study and field verification

26

Consequently, we have three scenarios for our business case analysis, to provide 3G mobile services in an isolated rural community:

a) “Satellite”: The mobile operator uses a satellite link to connect the access node (mobile base station) to the core network.

b) “Terrestrial”: The mobile operator builds a radio tower to connect the access node to the nearest node in a multi-hop terrestrial wireless network that also has a gateway (at the headend) with the core network.

c) “Shared”: The mobile operator leases space in an existing radio tower owned by a local

backhaul provider and installs the mobile access node at the same location.

2.5 Cost model

Each element of the network architecture (see section 1.3) implies a cost of investment (CAPEX) and operating costs (OPEX), as detailed in deliverables [TUCAN3G-D41] and [TUCAN3G-D51]. Below we give the typical values for the subsystems that comprise a mobile site (access network node): radio tower, solar PV system, (satellite or terrestrial) backhaul link and small cells (mobile base station). Radio tower

The most important cost in a “terrestrial” scenario is building the antenna tower structure, which may vary depending on the tower height and location as depicted in the following table:

(Prices in USD) Jungle (90m)

Highland (15m)

Steel tower structure 32617 1256 Steel flat bar 390 390 Lightning protection 977 626 Obstruction light LED solar 670 670 Construction materials 4428 779 Cables 632 142 Mounting accessories 99 95 Transportation 5600 1486 Building and installation 11800 5800 TOTAL 57213 11244

Table 11: Estimation of costs for antenna support structure Solar PV system

The major components for solar PV system are solar panels, charge controller, inverter, and batteries, although mounting, cabling and other electrical accessories are also needed to set-up a working system. Prices for PV systems have rapidly declined in recent years and depend strongly on the market, the size of the system, and used technology. In the United States, prices for utility-scale systems were around $0.90/W in 2012, while prices for smaller rooftop system in the highly-penetrated German market fell below €1.40 per watt in 2014. In that market, solar panels make up for 40 to 50 percent of the overall cost, leaving the rest to installation labour and to the PV system's remaining components. The table below shows an estimation of costs for the Tucan3G demonstration platform (terrestrial backhaul) depending on the nominal capacity (Cnom in Wh) of the batteries. The costs for the installation and transportation were already included in the estimation for the antenna towers above mentioned.

Page 27: TUCAN3G business model: case study and field verification

ICT-601102 STP Document number: D34 Title of deliverable: TUCAN3G business model: case study and field verification

TUCAN3G_D34EHASd.docx 27

Prices are in USD 1380 Wh 2760 Wh 5520 Wh Solar panels 210 420 840 Batteries 248 496 992 Regulator and accessories 196 196 196 TOTAL 654 1112 2028

Table 12: Estimation of costs for solar power system For a TDM/TDMA satellite link, power consumption is about 25-50W, which has a cost up to 6000 USD. Wireless terrestrial backhaul link

The costs of WiFi antennas and routers are negligible when compared to the radio tower costs. The table below shows an estimation of costs for the Tucan3G demonstration platform. The costs for the installation and transportation were already included in the estimation for the radio towers above mentioned. According to the different configurations described in the network design plan, the number of antennas and devices installed in each tower varies to meet the technical requirements, and this explains the differences between minimum and maximum costs.

Prices are in USD 1 antenna 2 antennas 3 antennas

Routers, switchers 292 394 496

Antennas 170 340 510 Cables and accessories 150 300 450

TOTAL 612 1034 1456

Table 13: Estimation of costs for wireless IP subsystem Satellite backhaul link

According to the analysis detailed in document D51, a professional TDM/MF-TDMA station sold as a VSAT kit can be around 2500 USD. Satellite kits for residential Internet access can be found for less than 1000 USD. In VSAT link installation, the main factors determining the OPEX are:

• Use of shared resources (not bandwidth) at the central Hub • Bandwidth (shared or not), according to service profile • On-site maintenance service and spares

Satellite bandwidth cost is high. Reference prices for LatAm region are 3500 USD MHz/month for Ku- Band and between 2800 (hemispherical coverage) and 3500 (global coverage) USD MHz/month for C-Band. With such an expensive resource, optimization of Bits/Hz ratio is a key driver. An example presented in deliverable D51 shows that OPEX in a VSAT link may change very significantly depending on the chosen configuration and the contention rate. Small cells (mobile access network subsystem)

The solution proposed in Tucan3G is based on outdoor femtocells (HNB) base stations to be installed in the antenna towers of the existing backhaul infrastructure. According to [SCF005], the wholesale cost for the consumer femtocells is between 100 USD and 300 USD. In the Tucan3G pilot, HNBs

Page 28: TUCAN3G business model: case study and field verification

28

provided by IP Access will be S-Class 16 with a cost of 770 USD including the additional equipment (container) for an outdoor deployment. The expected lifetime of a femtocell is less than for a macrocell, because they were originally designed for using indoor. Our analysis has been done for an estimated lifetime about 4-5 years. Rural outdoor locations using small cells have special installation requirements that may increase the cost. For outdoor mounting, depending on the environmental local conditions, special housing protection will be needed that indoor femtos do not. In addition, extended temperature range operation will probably need to be considered. Additional costs for the transportation, installation and configuration of the femtocells are much more significant than the cost of acquisition. Our estimation for the TUCAN3G pilot is about 2,150 USD in “Selva” and 1,950 USD in “Sierra”. The radio network controller (RNC) of the femtocells is installed at the headend where there is also a gateway link to the core network. In the TUCAN3G pilot, the cost of the RNC (model NC200 supplied by IP Access) was about 250000 USD. As said before, we assume the costs of RNC is prorated among 400 sites in the same network.

2.6 Social benefits (SROI)

In order to validate the cost-efficient technical solution proposed by TUCAN3G in a socially impactful way, it needs to be evaluated against a set of key performance indicators (KPIs). Short-term KPIs measure the immediate results of the activities – 3G mobile services access and use- whereas long-term KPIs are more related to the outcomes and social impact ultimate goals that TUCAN3G is pursuing -bridge the gap between urban and rural living conditions in developing countries.

Table 14: KPIs to measure 3G mobile services access and use

Objectives Indicators

Expand mobile coverage to isolated rural communities

- Number of communities served (at least one cell tower installed) - Number of inhabitants (population) within the coverage area - Number of public institutions (health, education, police, etc.) within

the coverage area

Provide access to internet using 3G mobile data service

- Available bandwidth (Mbps) for data connections per user

Increased adoption of ICTs in rural households

- Percentage of households with at least one mobile phone - Percentage of households with at least one smartphone - Percentage of households with a computer - Mobile telephone subscriptions per 100 inhabitants - Mobile internet subscriptions per 100 inhabitants

Increased use of ICTs in rural households

- Total number and duration of mobile telephone calls (voice traffic) managed per access network node

- Total number and volume of mobile internet connections (data traffic) managed per access network node

- Frequency of mobile telephone calls (originated and received) per user or household

- Frequency of mobile internet connections per user or household

Most of the short-term indicators can be measured by simply collecting traffic data from the network management and operation system. However, we found that it is not possible to obtain disaggregated mobile usage data per individual (mobile line) because it is against the privacy law. Such information can only be gathered by interviewing users. Likewise, those indicators related to the social impact need a field survey. Furthermore, long-term indicators need to be measured beyond the project

Page 29: TUCAN3G business model: case study and field verification

ICT-601102 STP Document number: D34 Title of deliverable: TUCAN3G business model: case study and field verification

TUCAN3G_D34EHASd.docx 29

lifetime. For instance, the reduction of poverty in terms of GNI per capita is an impact that will require at least one year being observable.

Table 15: KPIs to measure economic and social impact of 3G mobile services

Objectives Indicators

Generate enough revenue to achieve financial sustainability

- Total revenues generated per rural community - Average revenue per cell tower site - Average Revenue Per User (ARPU) - Frequency and amount of prepaid recharges

Increase productivity of local business

- Percentage of active users who have made mobile calls for productive purposes

- Percentage of active users who have accessed mobile internet for productive purposes

Improve social inclusion

- Percentage of active users who have made mobile calls for personal purposes

- Percentage of active users who have accessed mobile internet for personal purposes

Foster rural economic development

- Household disposable income - Rural poverty headcount ratio at national poverty line - Employment and employment-to-population ratio

What we need to measure in order to validate the business model hypothesis is basically the number of new mobile subscriptions and how are they using the voice and data services provided by TUCAN3G. The total revenues generated from the TUCAN3G pilot network will help us to validate the financial sustainability of the business model; the mobile operator billing systems can provide this data. However, the social impact is greater the more users there are. At equal total revenues, we prefer a 100% mobile phone penetration rate even if the ARPU is low rather than fewer users with higher ARPU.

2.7 Data sources

Field survey

The baseline survey was conducted in August 2015, we examined data on 488 family members (>= 12 years old) living in the four locations: Negro Urco (117), Tuta-Pischo (95), San Gabriel de Varadero (253) and San Juan (23). One of the four communities had 2G mobile coverage, San Gabriel de Varadero. We have analysed the adoption and usage of mobile services, even in those communities where there is no mobile coverage, because we found that some people use their mobile phone when they travel to the city (and they have coverage). However, the main objective in our analysis will be to determine the percentage of mobile users in San Gabriel de Varadero (the only community with coverage) in order to predict how many users there may be after deploying 3G services in the other three unconnected communities. Moreover, we have analysed their use of mobile services and how much do they spend in order to estimate the potential revenues for the mobile operator. In the final survey, conducted at the end of May 2016, we surveyed the same households as in the baseline in order to compare results and analyse the socio-economic impact of TUCAN3G pilot. However, we found that some of the households previously surveyed were not present (they have moved to other location) and sometimes some of the family members were no longer living in the same household.

Page 30: TUCAN3G business model: case study and field verification

30

Traffic data (during TUCAN3G pilot)

Traffic data was measured during TUCAN3G field trial period through the NOS management console provided by the NC200 network controller. Data provided was the number of calls and Erlangs of CS services (voice) and the Megabytes at PS (data) interface. We collected real data traffic measured by the femtocells in the TUCAN3G pilot during a two month period (19/Mar to 20/May). From these data, we estimate the total (and average) minutes of calls made by users located in the 6 rural communities targeted in the TUCAN3G pilot: Libertad, Negro Urco, San Luis de TacshaCuraray, TutaPishco, San Gabriel de Varadero and San Juan de Armanayacu.

Page 31: TUCAN3G business model: case study and field verification

ICT-601102 STP Document number: D34 Title of deliverable: TUCAN3G business model: case study and field verification

TUCAN3G_D34EHASd.docx 31

3 BUSINESS CASE ANALYSIS 3.1 Financial model

In this section, we present the financial model using the metrics defined in section 2.1 for each of the three scenarios defined in section 2.4:

• Scenario 1 (“satellite”): A location that needs investment to deploy new infrastructure (antenna support structure, solar PV system and telecommunications equipment) and integrated into a transport network via a satellite backhaul (VSAT) solution. Although this is not a deployment scenario in Tucan3G pilot, it is the most common situation for expanding mobile services to a new location in rural areas and will be used for comparison purposes.

• Scenario 2 (“terrestrial”): A location that needs investment to deploy new infrastructure (antenna support structure, solar PV system and telecommunications equipment) and integrated into the transport network via a radio link (WILD).

• Scenario 3 (“shared”): A location already integrated into an existing (shared) backhaul network that will need minor investment to upgrade existing infrastructure (deployment of femtocells for the mobile services). As the backhaul owner will provide the transport network service, the mobile operator will not need to deploy new costly antenna support structures.

Total cost of ownership (TCO)

First, we calculate the CAPEX and total OPEX for a period of 5 years, using the assumptions and cost model described in sections 2.2 and 2.5 respectively.

• For the “satellite” scenario, the access network node would require the deployment of one antenna supporting structure and a VSAT TDMA station. The solar PV system would supply the energy for the satellite station (about 50W). The integration with the transport network would be through the satellite link. All these costs, together with transportation and installation expenses, are included inside the concept “tower”.

• For the “terrestrial” scenario, in jungle the backhaul network would require a 75m antenna tower and WiFi router. The solar PV system capacity would be 1380 Wh. The cost of radio tower includes the transportation and installation.

• For the “shared” scenario, existing antenna support structures will be used to install the femtocells required for the access network. We will assume that the backhaul network also includes the connection to the core network. The costs included in the “tower” concept are limited in this case to the transportation and installation of the femtocells.

For all the scenarios, the cost of a femtocell would be the same and the cost of headend equipment (RNC – radio network controller) is calculated per node for a network of 400 nodes. With these hypothetical considerations, the estimated CAPEX would be the one shown in Table 16.

Prices are in USD Satellite Terrestrial Shared

Headend 625 625 625 Tower 11244 54087 Solar PV 6000 654 Backhaul link 1500 612 Small cells 2920 2920 2920

TOTAL 22289 58898 3545

Table 16: Estimation of CAPEX

Page 32: TUCAN3G business model: case study and field verification

32

Regarding operational costs, the most critical factor for the “satellite” scenario is the satellite link (VSAT service rate). As detailed in deliverable D51, the OPEX for a satellite solution mostly depends on the bandwidth requirements, and therefore, optimization of the Mbits/MHz ratio is a key driver. According to the technical requirements for the access network and the market reference prices for LatAm region presented in the deliverable D51, the estimated costs for a VSAT link (2Mbps DL/1Mbps UL and 1:2 contention rate) would be approximately 2,350 USD monthly, which means 28,200 USD in a year. For the shared scenario, as said before, we assume that the backhaul provider will charge a rent cost equal to 60% of the mobile operator revenues This is the worst case, since ideally there will be at least two tenants per radio tower. With these hypothetical considerations and the estimations detailed in D51, the operation and maintenance costs would be approximately a 10% of the CAPEX.

Prices are in USD per year Satellite Terrestrial Shared

Operation and maintenance 2229 5890 355 Backhaul service and tower lease 28200 7268

TOTAL 30420 5890 7622

Table 17: Estimation of OPEX per year.

In a period of 5 years, the estimated total costs (TCO) of the `terrestrial` scenario are much lower than the current solution based on satellite as is shown in Table 18.

Total in 5 years Satellite Terrestrial Shared

CAPEX 22289 58898 3545 OPEX 152145 29449 38110

TCO 174434 88347 41656

Table 18: Evolution of TCO for each scenario in 1- 5 years.

0

20000

40000

60000

80000

100000

120000

140000

160000

180000

200000

Y E A R 0 Y E A R 1 Y E A R 2 Y E A R 3 Y E A R 4 Y E A R 5

TCO (1-5 YEARS)

Satellite

Terrestrial

Shared

Page 33: TUCAN3G business model: case study and field verification

ICT-601102 STP Document number: D34 Title of deliverable: TUCAN3G business model: case study and field verification

TUCAN3G_D34EHASd.docx 33

There are clear differences in the costs of the three scenarios, which are estimated from the mobile operator point of view. In the short term, satellite has a lower total cost of ownership because it does require less capital investment (CAPEX). However, the operational costs (OPEX) related to the satellite link make it the most expensive option in the long term. After about 3 years, the terrestrial solution becomes a clear alternative for a mobile operator that wants to deploy its own (new) infrastructure. Obviously, sharing an existing infrastructure for the backhaul network is the lowest cost scenario in any case, because it does not require an initial capital investment (CAPEX) and the operational costs (OPEX) are reasonable low. From a cost analysis perspective, we conclude that a terrestrial wireless backhaul is a lower cost alternative to a satellite solution, although it would require an initial capital investment when there is no network infrastructure that could be shared. To evaluate if the investment is feasible we have now to calculate the ROI and the NPV for the terrestrial scenario, but in order to do so we previously need an estimation of revenues.

Revenue estimation

As said in section 2.2, a baseline field survey (performed before starting the TUCAN3G pilot) helped us to estimate the number of cell phone users and how much do they spend, since small isolated communities have presumably a lower ARPU than the country average. The first question to answer through the field survey was how many inhabitants have a cell phone and what are their characteristics?

• We found that 134 rural people in the data sample (27.5%) already have a cell phone, but only 48 out of them (9.8%) have a smartphone. It is noticeable that San Juan has the largest proportion, given that they did not have mobile coverage at the time we conducted the baseline survey. In our opinion, this is because this rural community is near Yurimaguas (about an hour away along a country road), which is the capital of Balsapuerto district. Also remarkable is the fact that the two other rural communities without mobile coverage, Negro Urco and Tutapishco, had 14.5% and 17.9% of mobile phone user penetration.

• Further analysis of bivariate distributions gives us more insights about the mobile phone users.

We wanted to examine the relation between personal characteristics (such as the role inside the family, age, gender or literacy) and mobile phone user penetration. Among family heads, 46.9% have a cellular phone. The proportion of households with at least one mobile phone user (probably owned by the family head) is greater than the percentage of individuals that have a mobile phone. The reason may be that most of the families share one cellular phone because they cannot afford having more, but we cannot confirm that without considering the income. Most of the family members (88.9%) earning more than $ 154 monthly have a cell phone, whereas only a 22.6% of those whose personal income is lower are mobile users. Moreover, a 100% of non-head family members (in the data sample) earning more than $ 123 monthly are mobile phone users.

Locality Cellphone Smartphone

San Juan de Armanayacu 56.5% 43.5% San Gabriel de Varadero 34.4% 11.5% Huitotos de Negro Urco 14.5% 1.7% Tutapishco 17.9% 17.9%

Total 27.5% 9.8%

Table 19: Marginal probability of having a cell phone and smartphone

Page 34: TUCAN3G business model: case study and field verification

34

• Among the people between 30 and 40 years old, the proportion of mobile phone users (40.8%) is greater than the marginal probability. On the other hand, people younger than 20 and older than 60 are less prone to have a cell phone.

• Among the women, the proportion of mobile phone users (17.9%) is less than the probability of men having a cell phone (36.3%), which is related with the fact that 93.1% of family heads are men. This difference is probably related with the generalized gender inequality, but a specific gender research should be performed on this issue to provide a deeper insight.

• Finally, the illiteracy is clearly a limiting factor for mobile phone user penetration. A vast majority (97.8%) of people who are illiterate do not have a cell phone.

The other important question was how often do rural people use their cellphone and how much do they spend?

• For this analysis, it is important to consider that Varadero was the only rural community in the baseline that had mobile coverage when the study was conducted. The mobile users living in the other localities need to travel to the city when they wanted to use the mobile service. As explained before, San Juan is quite near Yurimaguas, and therefore this fact explains why a 53.8% of mobile users who live in San Juan made at least one mobile phone call in the week before the survey. Among the cell phone users in Varadero, a 46% of them make phone calls in relation with both their productive activity and personal issues, whereas a 54% only make personal calls to talk with their family and friends.

Table 20: Frequency and reason of mobile users making phone calls

Locality

Frequency Reason

Daily Weekly Personal Productive and Personal

San Juan de Armanayacu 23.1% 53.8% 61.5% 38.5% San Gabriel de Varadero 41.4% 51.7% 54.0% 46.0%

• Among the mobile users in Varadero, the vast majority (72.4%) do not use the mobile internet

service, only a 5.7% use it for personal purposes whereas a 21.8% use mobile internet for their productive activity. Here it is important to note that data connectivity was based on GPRS at that moment, and consequently Internet access was slower than nowadays with 3G. This can influence the use of mobile internet services from now on.

• For the analysis of mobile expenses using baseline data, we only consider the mobile users in

Varadero, since the other rural communities have strong limitations in the supply of mobile services (no coverage) as explained before. Obviously, this fact has an influence on how frequently and how much people living in San Juan, Negro Urco or Tutapischo spend in mobile services. Among the mobile users in Varadero, the average expenditure in the week before the baseline study was $ 2.67, but 50% of them spent between $ 0.31–2.77. Those users that made mobile phone calls and access internet through their mobile phone (using both voice and data services) spend twice more on mobile services. If we only consider internet mobile users, then the average expenditure was $ 5.21 in a week.

• Likewise, mobile users earning a personal income (including family head) have a higher

mobile expenses average ($ 3.17). However, there is no lineal relationship between personal (or household) income and mobile expenses. The correlation factor was under 0.4, we could not predict mobile expenditure based on the income with enough accuracy.

In the TUCAN3G pilot, the estimates of total revenue were validated by comparing it with the data collected from real traffic monitoring reports. The network management system (NOS Client) provides some KPIs to measure the performance of the femtocells: number of voice call, total voice

Page 35: TUCAN3G business model: case study and field verification

ICT-601102 STP Document number: D34 Title of deliverable: TUCAN3G business model: case study and field verification

TUCAN3G_D34EHASd.docx 35

minutes, number of data connections and data traffic (in MB). If we use this data to make an estimation of revenues in each location, we obtain the actual revenues during the TUCAN3G pilot. Figure 7 shows the number of voice calls per week during the last two months, which help to illustrate how much they use the service in each community.

Figure 7: Voice traffic measured during TUCAN3G pilot

In order to calculate the voice revenue using data provided by the NOS Client (Table 21), we multiply the total voice calls per the average call duration (in minutes) and per the voice ARPM (Average Revenue Per Minute). The value of ARPM used ($ 0.05) was obtained after consulting with MNOs, and is an average value for the combination of two different categories: incoming and outgoing calls, and prepaid and post-paid services.

Table 21: Voice revenue measured during TUCAN3G pilot. # Calls

weekly Average call duration

(minutes) Annual

revenue ($) Population #

Users ARPU

($) San Juan 300 1.381 1512 47 13 10 Libertad 1296 2.337 11054 331 89 10 Negro Urco 1965 1.943 13936 383 103 11 TutaPishco 1175 2.006 8604 305 82 9

Average 1104 1.999 8349 251 68 10 Then we estimate the revenue generated from data traffic using a similar method: we calculate the volume of data (45,947Mbits for the hole network) and multiply by the average revenue per Mbit ($ 0.015, provided by MNOs), in order to estimate the revenue of the data service ($ 689 annually per location) and the data ARPU ($ 1 per user in a month). Last, we add both revenue streams -voice and data- to get the total revenue per user (ARPU monthly), getting $ 11. For this calculations we didn’t use data from TacshaCuraray and Varadero because at those communities 3G and 2G networks coexist, and we don’t have information about how much traffic is going through the 2G network, so we can’t estimate the whole revenue at that locations. In fact, we know that revenues in Varadero before TUCAN3G (with the 2G network only) doubled current revenues in the 3G network. Therefore, so we can assume that half of the traffic is going through the 2G network, because the use of cell phone is probably the same. Moreover, when comparing 2G ARPU in Varadero before TUCAN3G pilot with current 3G ARPU in the rest of communities, we have found that they are very similar. This

0

500

1000

1500

2000

2500

3000

19/03 al25/03

26/03 al01/04

02/04 al08/04

09/04 al15/04

16/04 al22/04

23/04 al29/04

30/04 al06/05

Voice calls per week

Varadero San Juan Tacsha Curaray

Libertad Negro Urco Tuta Pishco

Page 36: TUCAN3G business model: case study and field verification

36

implies that demand in small jungle communities follows a similar pattern, and then we can assume that removing Varadero and TachsaCuraray from our calculations has a very low impact on our results. Return on Investment (ROI) and Net Present Value (NPV)

We can now calculate the ROI and the NPV for the TUCAN3G pilot, as shown in Table 22. First, we calculate the annual ROI for the first 5 years (as the gross profit divided by the CAPEX) for the average case of TUCAN3G locations (91 users, an ARPU of $ 11 monthly, with 5% revenue growth) according to the estimated costs described before. Then, we obtain the net cash flows for each year in the period considered. The discounted cash flows minus the initial investment give the Net Present Value, using the formula described in section 2.1. As said before, should that value be positive for a given n year period, it means that we are creating value and the project will be profitable. The value of such n is the Discounted Payback Period. We calculate the NPV at the end of the first 5 years assuming a subsidy equal to the costs of building the radio tower structure and backhaul network equipment. With those assumptions (subsidies), the NPV becomes positive in the first year (Table 22). Therefore, the payback period is under 1 year considering a discount rate of 9%, which is an attractive scenario for a MNO. Later, in the chapter “Sensitivity analysis”, we will analyse what happens to the payback period when there are no subsidies to cover the costs of building the radio towers, and will see that it may become negative depending on factors such as number of users, percentage of subsidies received or geography conditions.

Year 0 Year 1 Year 2 Year 3 Year 4 Year 5 Subsidies 55.353 Network equipment 58.898 Revenue 12.055 12.657 13.290 13.955 14.652 Network oper. &maint. 5.890 5.890 5.890 5.890 5.890 Gross profit 6.165 6.768 7.400 8.065 8.763 ROI 10% 11% 13% 14% 15% SGA (indirect) costs 603 633 665 698 733 Depreciation 3.164 3.164 3.164 3.164 3.164 Operating Profit 2.398 2.970 3.572 4.203 4.866 Operating profit margin (%) 20% 23% 27% 30% 33% Tax 671 832 1.000 1.177 1.362 Profit 1.726 2.139 2.572 3.026 3.503 Net profit margin (%) 14% 17% 19% 22% 24% Net cash flow -3.545 4.891 5.303 5.736 6.190 6.668

Net Present Value (NPV)

942 5.405 9.835 14.220 18.553

Table 22: Net cash flow calculation with subsidies

3.2 Analysis of results

As explained in the previous sections of this chapter, many network design choices will determine the final costs. For instance, the satellite link may be SCPC or the cheaper TDMA alternative; the radio links in the backhaul network can use WIMAX or WILD; and the number of antennas and telecom equipment varies according to the bandwidth requirements. However, the most critical factor driving costs (CAPEX) is the height of the antenna support structure. When there are no existing radio towers (a new deployment), total costs of investment for a wireless backhaul are quite significant, much more in the jungle than in the mountain areas. As said before, radio towers in the jungle could require about 90m, whereas in mountain areas a 15m tower is enough. In our financial analysis, we considered a site

Page 37: TUCAN3G business model: case study and field verification

ICT-601102 STP Document number: D34 Title of deliverable: TUCAN3G business model: case study and field verification

TUCAN3G_D34EHASd.docx 37

location in the jungle, which is the worst case. The financial results would still be positive in the highland rural communities, because the cost of building a new radio tower is much lower. However, new deployments in jungle communities may require subsidies for building radio towers, which will depend on the expected number of users (or total population) in the area to be covered. Despite the lower total costs of the dedicated terrestrial wireless network, moving OPEX to CAPEX could be a shortcoming for a mobile operator. Both the shared and satellite cases have a similar operational costs structure in terms of CAPEX-to-OPEX ratio. Therefore, a shared rural network used for backhauling provides additional value for the MNO to lower their total cost of deployment in isolated rural areas. A decision to build or lease will ultimately depend on the expected revenues and the negotiation of the lease contract for the use of the backhaul network. Sharing an existing infrastructure for the backhaul network is the lowest cost scenario in any period, because it does not require an initial capital investment (CAPEX) and the operational costs (OPEX) are significantly lower. Considering these cost savings, the TUCAN3G business model would enable sustainable business operations (for the MNO) in population centres with at least about 100 mobile phone users. Furthermore, a rural backhaul provider sharing the existing network infrastructure for new commercial services will generate revenue from the MNO, which will contribute into leveraging the costs of new network deployments. The lease rate should cover at least the OPEX for the backhaul network only with only one tender, and the low expected revenues for the MNO would determine the maximum lease rate in rural communities with a very low population density.

3.3 Non-financial results

Based on the field survey analysis we have some preliminary results of the socio-economic impact after the field trial. Despite the duration of TUCAN3G pilot has not been long enough to show any important change of life conditions, we can appreciate some evidence of positive social impact. The comparison of results before and after the field trial is summarized at Table 23. In the short period of time that the TUCAN3G pilot lasted, we could not increase the number of mobile users, but we appreciate 0.5% more smartphone users. The most remarkable data is that now 51.5% of mobile phone users receive at least one call per day and 44.3% make a call daily. That is an increase of 26.1% and 15.2% respectively in the use of mobile phones for voice calling. However, we do not appreciate a similar enthusiasm in the use of mobile internet, just a 6.5% increase although 5.5% more users said that mobile internet is useful for their productive activity. The mobile expense, as reported by the users surveyed (as opposed to revenue estimated from data traffic), has increased a 47.8% (mean value). The median value of mobile expenses per user is now about $ 6 monthly and the mean is around $ 11, what matches with the revenues estimation based on actual traffic from the pilot. Comparing this data with our forecast, we are still too far from the ITU-D indicators for developing countries (90% mobile telephony and 40% mobile broadband). However, we found that the average revenue “per inhabitant” ($ 33) is quite similar for the locations without previous mobile coverage. Moreover, this estimate is quite similar to the average revenue per inhabitant for the 2G mobile services in San Gabriel de Varadero (before TUCAN3G pilot).

Page 38: TUCAN3G business model: case study and field verification

38

Before After Diff. Mobile users (as % of total population) 27.5% 24.9% -2.6 % Smartphone owners (as % of total population) 9.8% 10.3% 0.5 % Call making frequency (as % of mobile users)

daily 29.1% 44.3% 15.2 % weekly 38.8 % 32.0 % -6.8 % less frequent 32.1 % 23.7 % -8.4 %

Call receiving frequency (as % of mobile users) daily 25.4% 51.5% 26.1% weekly 41.8% 34.0% -7.8 % less frequent 32.8 % 14.5 % -18.3 %

Use of mobile internet (as % of mobile users) business 17.16% 22.7 % 5.5 % personal 5.22% 6.2 % 1.0 % no use 77.62 % 71.1 % -6.5 %

Expenses in mobile use (in $ monthly) Mean 7.59 11.22 47.8% Median 3.64 6.06 66.7%

Expenses or mobile internet users (in $ monthly) Mean 10.91 15.65 43.4% Median 10.91 9.09 -16.7%

Table 23: Results of the field survey before and after the field trial

Page 39: TUCAN3G business model: case study and field verification

ICT-601102 STP Document number: D34 Title of deliverable: TUCAN3G business model: case study and field verification

TUCAN3G_D34EHASd.docx 39

4 SENSITIVITY AND RISKS ANALYSIS 4.1 Sensitivity analysis

Subsidies for new deployments

Obviously, the estimate presented in section 3.1 would not have been so profitable if there was no existing infrastructure (or subsidies for building radio towers). For new deployments, the TUCAN3G business model proposes that a public or private institution owns or subsidises the rural infrastructure, which can be used for backhauling, providing fixed services and supporting public institutions such health, education or government establishments. That is, a multi-stakeholder partnership between the mobile operators that provides services to end-users, the rural network operator that manages the infrastructure and public/private institutions providing funding to build a rural backhaul network (Figure 8).

Figure 8: The TUCAN3G multi-stakeholder partnership model

Based on the financial model presented in section 3.1, we analysed what happens if the assumption about subsidies for the radio tower building is not held. For this matter, we suppose a case with 100 users (what means 370 inhabitants with a 27% penetration), an ARPU $11 monthly and 5% annual growth in revenues (Figure 9). We start (axis origin) without subsidies and see what happens to the payback period compared to the desirable reference of 5 years. We see that 70% or more subsidies are needed for a payback period shorter than 5 years in the jungle areas, whereas highlands locations do not need any subsidies.

Figure 9: Sensitivity of payback period to subsidies

Page 40: TUCAN3G business model: case study and field verification

40

Another way to achieve better financial results would be to increase revenue. Since this mostly depends on the number of users per location, let us see what happens to the payback period if we increase the number of users. Again, the axis origin represents the reference case of a 5 years payback period with ARPU $11 monthly and 5% CAGR. Now, we see that a total number of 200 users is needed for a 5 years payback period without subsidies (Figure 11) or 140 with a 40% of subsidies (Figure 11). In the first case, with only 100 users the payback period in the jungle areas is much longer than 5 years, which is not acceptable for an investment by a mobile operator. Noticeable is that highland locations have a 3 years payback period even with only 80 users.

Figure 10: Sensitivity of payback period to number of users without subsidies

Figure 11: Sensitivity of payback period to number of users with subsidies

Finally, we analyse what combination of users and subsidies is needed to have a payback period of 5 years, in the same conditions. As we see in Figure 12, a 100% of subsidies will guarantee that it is profitable to provide 3G mobile services in jungle communities with 50 users (185 inhabitants).

Page 41: TUCAN3G business model: case study and field verification

ICT-601102 STP Document number: D34 Title of deliverable: TUCAN3G business model: case study and field verification

TUCAN3G_D34EHASd.docx 41

However, without any subsidies, at least 190 users (700 inhabitants approximately) are required to have a payback period of 5 years. For highland locations, subsidies are needed if the number of users is below 50 (about 185 inhabitants with a 27% mobile user penetration).

Figure 12: Subsidies and users needed for a 5 years payback period Subsidies for new deployments

We have just seen that, despite TUCAN3G manages to reduce TCO in 5 years; subsidies will still be required to reach communities smaller than 1.000 inhabitants. In order to understand where these subsidies could come from, in this section we review some public and private initiatives that could help to expand connectivity to isolated rural communities. The big Internet companies – Facebook, Google and Microsoft – have plans to support broadband development in Colombia and Peru:

• Recently, Facebook’s founder Mark Zuckerberg announced his plan to bring free internet to Colombia. With Colombia's tech industry growing 177% to $6.8 billion according to official statistics, with 46 entrepreneurs graduating and 105 start-ups emerging from the ‘HubBOG’, the 5-year-old program for tech entrepreneurs, Colombia is becoming the “Silicon Valley of South America”7. In Chocó, which is Colombia’s department where most social inequality and lower living standards exist according to the United Nations, Microsoft invested in software called ‘Healthicloud’. Healthicloud helps filing information about patients’ health conditions and test results. After the successful involvement of Microsoft in the health services provided by the local clinic called ‘Life Clinic’, Microsoft expanded its investment with its project including new partnerships with the cooperative ‘Médicos Especialistas del

7http://money.cnn.com/2015/03/13/investing/colombia-tech-silicon-valley/.

Page 42: TUCAN3G business model: case study and field verification

42

Chocó’ and the local hospital of ‘Clínica Virgen María’8. In the same underserved area of Chocó, Microsoft invested in education, by providing tablets and laptops with its own software of Windows and Office to young students of the ‘NuestraSeñora de la Pobreza’ school9.

• In Peru, in the rural areas of Cajamarca, Andahuyillas, Ayacucho and Cusco, Fundacion

Telefonica has taken initiatives to use the internet and computer technology for education purposes; they invested in “Mobile Classrooms” which provide usage of tablets, internet connection and computer programs to approximately 25,000 Peruvian students.

• In Colombia, there are some enterprises specialized in providing services to different sectors

as oil companies, construction, agro industrial, offering services such as internet, voice and data transmission, allocated solutions SCPC, continuity to backup and backhaul business. Even when these enterprises can be competency for TUCAN3G, it is possible to consider an alliance for using its infrastructure with the objective of extending mobile coverage to rural areas. Some of this enterprises are:

• AXESAT was founded in 2003 by Colombian businesspersons as a solution to provide

internet connection for business operations in places with poor connectivity. Their vision is to become the best provider of satellite services for Latin America enterprises. They have installed more than 11.000 active links with support from technological partners such as Gilat, iDirect, Hughes, Intelsat, Eutelsat and SES.

• Azteca Comunicaciones Colombia was the enterprise selected in 2011 by the national

government to design, install, and manage the operation and maintenance of the national optical fiber infrastructure expansion, within the project named Proyecto Nacional de Fibra Óptica, in approximately 753 municipalities and 2.000 public institutions. This enterprise also offers services to SMEs, cooperatives and individual customers. Azteca Comunicaciones has deployed network nodes across the national territory to provide services such as, co-location (site sharing) for customers who require increasing their telecommunication services without making high investments.

Besides the private enterprises, initiatives in Mobile for Development, there are also public funding opportunities in what is called the Universal Access. The Universal Access and Service (UAS) funds aim at stimulating network deployments in rural and remote areas that are not served or under-served, with particular focus on low-population density areas where provision of services is not normally economically viable. Policy-makers tackle the availability problem by subsidizing the deployment of network infrastructures that interconnect remote and rural areas because they believe that pushing broadband supply will bring greater adoption. As consequence of high mobile penetration, policy-makers are stimulating mobile broadband access in their rural development strategies. Public subsidy policies for digital inclusion are driving competition and the modernization of wireless network infrastructures.

• In Chile, one of the objectives for 2020 in the "Plan Nacional de Infraestructura" is the deployment of a regional backbone "Fibra Óptica Austral" to provide high-capacity transport and interconnect the Magellanic people with the rest of Chile. The state will finance, through public tender, almost 100% of the installation costs of the project whose implementation should take place between 2016 and 2017. Thanks to the project "Todo Chile Comunicado", subsidized by Fondo de Desarrollo de las Telecomunicaciones (FDT), 1,474 rural villages were provided 3G mobile broadband by 2012, benefiting more than 3 million people.

8http://www.microsoftcolombia.com/Healthicloud_en_clinicas_choco (in Spanish) 9http://www.microsoftcolombia.com/Computadores-Para-Educar

Page 43: TUCAN3G business model: case study and field verification

ICT-601102 STP Document number: D34 Title of deliverable: TUCAN3G business model: case study and field verification

TUCAN3G_D34EHASd.docx 43

• Mexico recently launched an ambitious project to achieve full coverage in its rural areas where private operators have no incentive to invest. The "Red Compartida" is a single wholesale network using the 700MHz band released after the "analogue blackout" (digital dividend). The investment will be implemented through a Public Private Partnership with a total cost of about 130 million USD, of which the Mexican government would provide about 13 million. Although a consortium will operate it, the shared network is at the service of all mobile operators, including MVNOs. Thus, the new operators are helped to compete with the incumbent as they can improve their coverage without having to bear the cost to deploy a network.

• Regional broadband projects sponsored by the government of Peru will provide all district

capitals with the full capacity of a new fibre optic backbone (Red Dorsal Nacional de Fibra Óptica). FITEL will manage the investment's 21 projects, which will be tendered until 2017. The total investment amounts to 1162.3 million USD, of which 628.4 million will be invested in the access network (599.6 million will be subsidized) and 534.8 million in the transport network. This broadband infrastructure will provide high-speed Internet to 6,411 rural locations and benefit about 5 million Peruvians.

• In Colombia, the government is promoting the programme “Vive Digital” since 2010, aim at

promoting ICT use in order to reduce poverty and create employment opportunities. This plan is structures in four components: infrastructure, services, applications and users. The expansion of infrastructure is based on the implementation of the Optical Fibre National Project. Another project called “Proyecto de Conectividad de Alta Velocidad” will provide connectivity in 27 main towns and 20 small towns benefitting approximately 462,658 people living in the Colombian jungle.

Low-income people living in rural communities usually cannot afford mobile broadband services, even if they are available, because smartphones or tablets and the tariffs (broadband and mobile Internet plans) are not affordable for their level of expenses.

• In Chile, a new regulatory reform is expected to broaden FDT's scope by creating new lines of funding, subsidizing digital equipment, investing in digital literacy and content/application development and improving telecommunications infrastructures. The public Internet service “Zona WiFiChileGob" aims to reduce the digital divide in the most vulnerable places of Chile. It provides Chileans who have access to personal computers, mobiles, or tablets with 30 free minutes of WiFi. "Conectividad para la Educación" has been an inter-ministerial joint initiative with the goal of providing quality Internet access to more than 8,500 schools.

• Affordability is still a problem for many people in Mexico even though prices have decreased

by 21% between 2007 and 2010. "Mexico Conectado" promotes the deployment of telecommunications networks that provide connectivity in public spaces such as schools, health centres, libraries, community centres or parks. There are already more than 65.000 public sites with free broadband and 35,000 additional sites were deployed in 2015.

• In Peru, regional projects for the access network planned by FITEL will allow the provision of

free internet and intranet broadband for nearly 3,000 public institutions (school buildings, health facilities and police stations). The deployment of a National Fibre Optic Network (RDNFO) will connect 180 district capitals, a total investment of $ 333 million. The 21 Regional Projects planned will leverage this national backbone infrastructure to bring high-speed (2-4 Mpbs) internet access to more than 6.000 localities and benefiting more about 4 million inhabitants, a total investment of $1,800 million. The 18 new projects formulated by FITEL during 2015 amounted $703 million. Among them, there is one project subsidized by FONIE (Fondo para la Inclusión Económica en Zonas Rurales) which will bring mobile

Page 44: TUCAN3G business model: case study and field verification

44

services to at least 200,000 inhabitants living in 771 localities in the poorest districts of Peru, a total investment of $64 million.

Figure 13: FONIE Districts, Source: Dirección General de Gestión de Usuarios - MIDIS.

• The other components of the “Plan Vive Digital” in Colombia include the implementation of

953 “Kioscos Vive Digital” in rural areas and 56 “Puntos Vive Digital” for the poor people who cannot afford to buy a computer or mobile phone to access internet. FONTIC (Fondo de Tecnologías de la Información y Comunicaciones) is a public fund providing finance to social inclusive initiatives.

4.2 Regulatory issues

Public-Private Partnerships can play an important role in fostering long-term investment and economic growth for the rural poor. However, it is necessary to review regulation conditions that may influence investment in isolated rural areas. Among the issues that telecoms regulators traditionally have to deal with, those with the greatest impact on mobile operator's business are licensing, spectrum management and supervision of quality of service. In this section we analyse the regulation concerning these aspects in four countries of Latin America: Peru, Colombia, Chile and Mexico. We have selected these four countries because they apply complementary strategies that serve to enrich the regulatory analysis. Licensing and Authorization The number of mobile operators competing in isolated rural communities is less by necessity than in cities. If an operator already has deployed its network infrastructure, they rarely want to risk another operator investing in new infrastructure. Earnings expectations in the areas of low incomes are too low to share the profits with other operators and achieve a return on investment. By regulating the market entry of MVNOs, politicians can boost infrastructure sharing in order to increase the commercial offer, without needing to deploy new networks. Obviously, the MVNOs depend on the infrastructure

Page 45: TUCAN3G business model: case study and field verification

ICT-601102 STP Document number: D34 Title of deliverable: TUCAN3G business model: case study and field verification

TUCAN3G_D34EHASd.docx 45

deployed by the MNOs sharing it. Since the regulator determines the wholesale prices, the tariff plans offered to end users can be cheaper (small companies have lower fixed costs) and this may be threat for the incumbent mobile operator. Regulators can foster a competitive market by increasing the number of suppliers licensed to operate. In Chile, the number of mobile operators has increased, from three to eleven competitors, due to the entry of new operators with network infrastructures, such as are VTR Móvil (today is another MVNO) and WOM (before called Nextel), and the incorporation of new mobile virtual network operators (MVNO), such as Virgin Mobile, GtdMóvil, Netline Mobile, Falabella, Telstar and Simple. Following the merger of Nextel with Iusacell in Mexico, there are now three mobile operators with their own network, the other two being Telcel and Movistar. Telecommunications reform opened the window of opportunity for the arrival of mobile virtual network operators. However, there is still no specific legislation to define the role and requirements for such operators. The “Instituto Federal de Telecomunicaciones (IFT)” has opened a public consultation to define aspects such as the responsibility of quality of service and customer care. It is expected that the new shared wholesale mobile network will foster the launch of new MVNOs. At the end of 2014, Peru approved the new telecom law No. 3008310 to define the role of MVNOs. Movistar, Claro, Entel and Bitel are the four mobile network operators in Peru. Several companies have shown their interest in being an MVNO, such as Falabella and Virgin. In addition, that Peruvian new law defined the new role Mobile Rural Infrastructure Operator (MRIO). This role is only applicable to rural areas where mobile network operators have not deployed their own infrastructure yet. The MRIO provides the transport and mobile access network infrastructure in a rural area, but it has no end users. They must negotiate with a mobile operator to use their numbering and allocation of radio spectrum to provide public mobile services. At the same time, the MNO is obliged to use the rural infrastructure of the MRIO unless they have already deployed their own mobile access network. When it comes to sharing network infrastructures, it is important to define the responsibilities of the network operator (MNOs) versus the MVNOs that sells services to end customers. In the case of Mexico - Red Compartida - and Peru with the new role of MRIO, the regulator needs to clarify the matter. Portability, which allows citizens to retain their telephone number when changing their service provider, is another regulatory measure that promotes competition by preventing operators impose exit barriers to their customers. In Chile, the mobile portability rate in 2015 is estimated at 6.5%. As of June, the company has lost more mobile numbers is Entel (374.243). On the contrary, Claro has won mobile numbers 329.975 and 238.637 Virgin Mobile. Portability rules issued in November 2014 by the IFT, in Mexico, allow users to change their fixed or mobile telephone company without losing the number within a maximum of 24 hours. Entel Peru has been the winner in mobile number portability, earning 101.797 users in the period between July 2014 and February 2015, according to statistics report from national regulator Osiptel. One of the usual concerns to extend mobile coverage is the difficulty to obtain permissions to deploy new base stations, which is mainly due to: a) the discretionality of local authorities when defining requirements to build cell towers, and b) the general belief on the health risks posed by mobile antennas for people living near cell towers [Fedesarrollo13]. Most of local organizations, covered up by health prevention mechanisms, have established severe restrictions to build cell towers in residential zones, near to hospitals and schools. However, the World Health Organization has underscored in several times, making sure that until now, there is not any proof of a negative effect related to health caused by the cell towers. For example in Colombia, according to the ANE (National Spectrum Agency), there are 17 cities, where the situation is complicated because of local restrictions

10 Ley No. 30083: “Ley que establece medidas para fortalecer la competencia en el mercado de los servicios públicos móviles”

Page 46: TUCAN3G business model: case study and field verification

46

to install cell towers: Itagüí, Bogotá, Cali, Barranquilla, Cartagena, Santa Marta, Cúcuta, Ibagué, Villavicencio, Valledupar, Montería, Palmira, Barrancabermeja, Pasto, Neiva, Armenia and Soledad (Atlántico). Spectrum management The so-called digital dividend, analogue TV frequency release, is a unique opportunity to significantly increase radio spectrum allocated to mobile services. Because of the technical characteristics of these frequency bands, they are ideal for extending rural coverage. This could imply the entry of new competitors in the mobile market, although in the case of Mexico the regulator has preferred to take the opportunity to create a single shared wholesale mobile network. Chile will probably be the first country in Latin America to have LTE-Advanced mobile technology; soon there will be a second spectrum band for 4G - 700 MHz -, in addition to the current in use (2,600 MHz). In June 2014, ProInversion (in Peru) was commissioned to conduct the license auction for three blocks of spectrum in the 700 MHz band (known as the Digital Dividend) with the objective to popularize mobile Internet services using high-speed 4G LTE. Radio spectrum has a significant economic value that governments can leverage in order to obtain additional benefits (to usage fees) and meet their strategic goals for digital inclusion. For instance, in Chile, as consideration for granting use of the 2,6 GHz spectrum band (4G mobile services), Movistar, Claro and Entel must offer internet access in 543 isolated locations throughout the country. The regulation started operations in March 2015 and has benefited 27 localities in extremely remote areas. Peru’s government, after a long negotiation to renew the licensing of Telefonica in 2013, achieved its commitment to establish a social tariff, expand network coverage and offer free Internet in public institutions located in the isolated jungle rural communities. The Peruvian supervisor agency (OSIPTEL) determines the price for social tariffs. This applies to mobile calls made by users enrolled in social programs like Juntos, Cuna Más, Pensión 65 y Beca 18. The extension of rural coverage obligation implies the provision of mobile service in 100% of rural villages with more than 400 inhabitants. Public radio spectrum tender held in Colombia in mid-2013 included conditions to improve coverage of mobile services in rural areas. Licensees must ensure service provision in at least 29 locations, which are especially districts and villages located in rural areas of the 32 departments of Colombia (listed in resolution 000 449 of 11 March 2013). Some of these locations are in remote regions such as the Amazon where 66 localities and Orinoco regions where there are 128 localities were included. Moreover, the Colombian Ministry of Information and Communication Technologies (MinTIC) imposes minimum coverage obligations to mobile operators. In the municipalities located in rural areas, the required minimum mobile coverage is 30% except in the case of departmental capitals in which 50% is required. Quality of service National regulators must follow the recommendations of the global standardization of telecommunications systems and services, including QoS issues, published by the ITU agency. Mobile operators take on a commitment to quality of service when they are granted regulatory approval. The public body in charge of supervision is responsible for carrying out the monitoring of network performance indicators to verify that the quality exceeds the minimum limit established. A QoS approach oriented to penalties can discourage investment and reduce the supply of services, especially in rural areas where no electricity network and the use of solar energy is not as reliable. There are three types of indicators to measure quality of service (QoS), relating to: user complaints, coverage and capacity of the network, and user satisfaction. In general, required minimum limit values exist in the case of telephony, but not for mobile broadband. Isolated rural areas lack of reliable energy provision infrastructures makes the mobile operators more reluctant to provide their services, because the power blackouts could imply penalties for the consequent unavailability of mobile service.

Page 47: TUCAN3G business model: case study and field verification

ICT-601102 STP Document number: D34 Title of deliverable: TUCAN3G business model: case study and field verification

TUCAN3G_D34EHASd.docx 47

Mobile operators demand a more relaxed regulation when providing services in remote areas without electricity and regulators understand their concern. The Chilean regulatory changes in 2010 on the measurement of quality of service in mobile telephony sets different set of indicators for mobile base stations located in areas classified as rural. The percentage of call attempts and call completed successfully must be above 97% in urban areas and 90% in rural. In Mexico, IFT publishes a quarterly report including mobile internet performance measures such as failed and interrupted FTP sessions ratio, latency time and service times in FTP session, average speed in data download. The supervisor authority in Peru, OSIPTEL, has defined similar indicators for internet public services: minimum speed, average speed, link occupancy rate, data transfer rate, packet loss, and latency. In Colombia, the institutions in charge of supervision of service quality are Comisión de Regulación de Comunicaciones (CRC), Superintendencia de Industria y Comercio (SIC) and MinTIC. There have been established two zones: Zone 1 that includes those departments whose population represents at least 1% from national territory and Zone 2 composed by that other with less than 1%. The percentage of dropped calls for Zone 1 may not exceed 2% and the maximum allowed for Zone 2 is 5%. The supervisor always has the authority to order the implementation of precautionary and/or corrective measures that will ensure compliance with the regulation governing the provision and use of public telecommunications services. Another way to measure the quality of service, in a way, is through user satisfaction surveys. The supervisor also monitors and publishes statistics on the attention and resolution of user complaints filed against the operators. During 2014, OSIPTEL managed 3,640 cases that were managed satisfactorily solved for users and companies. In Mexico, during 2013, VTR was the company with the highest number of complaints, with a rate of 5 points. Claro (2.1 index points) followed it. Meanwhile, Movistar is the company that submitted the best ratio (0.7 points), i.e., a claim for every 10 thousand users.

4.3 Free public services

Besides the regulatory issues above mentioned, public policies encouraging the use of alternative telecom services can affect the value proposition of mobile operators. Public internet access in schools, health centres, community centres and public parks is a substitute service of mobile broadband. Given that most modern mobile phones can use Wi-Fi, free internet access available in these public places refrain users from spending their money on mobile broadband plans and this can be a serious threat to the mobile operators if there is no limit. In Chile, the users of the free wireless internet service - Zona WifiGob - have a use limit of 30 minutes per session. There is more freedom in the case of Mexico and Peru's free internet access, although the coverage is limited to some specific locations (small areas inside buildings and public places) and not everywhere, as when using mobile broadband. Given the lower speeds of mobile broadband when compared to internet access services based on fibre optic or ADSL, mobile broadband has a true competitive advantage in those very isolated rural communities where there are no plans to deploy high capacity transport networks. In Colombia, the government strategy to connect rural areas has been focused during the last years on the installation and operation of “Kioskos Vive Digital”11, which are public centres offering telephone and internet services to promote the use of ICTs and bring their benefits in economic, social, cultural and recreational activities to the poorest population. The “Kioskos Vive Digital” are located in underserved rural communities with more than a 100 inhabitants. They are mainly installed in schools but the community in general can use them up to 20 hours per week. Wi-Fi is available also with a range of 50m around the school for public free use. In the second phase of the programme, they were installed also in indigenous communities and they can be used up to 40 hours per week.

11 http://www.mintic.gov.co/portal/vivedigital/612/w3-propertyvalue-7059.html

Page 48: TUCAN3G business model: case study and field verification

48

4.4 Other public policies

Public subsidies and financial facilities for purchasing mobile devices to access broadband services may help to eliminate barriers and boost demand. The Chilean public fund FDT is considering the possibility of subsidizing the purchase of technological equipment by end users in order to encourage the use of broadband. Subsidizing the sale of mobile devices is a good strategy for the MNO, because it is reflected in the increase in data services demand. However, the current trend is to replace user terminal subsidies by micro-credits and other payment facilities. An interesting case for rural poor communities is Grameen Telecom in Bangladesh12, where the "phone ladies" provide cellular pay phone service at an affordable price to extend telecommunications to rural villages, where people cannot afford having their own device and other access alternative is not available. The mobile operators can provide access to energy for subscribers that live off-grid. The “GSMA Mobile for Development Utilities” initiative [GSMA13] has explored synergies between mobile, energy, water and sanitation services. The GSMA sees an opportunity for the mobile industry to enrich their value proposition, leveraging their solar energy systems in underserved isolated rural areas, and become Energy Service Company (ESCO) off-grid as Movistar has done in Nicaragua. Regulators should be flexible, allow innovative business models and encourage partnerships between NGOs and utilities. User training to promote the use of internet is necessary to overcome digital illiteracy among the rural. Publically sponsored initiatives like México Conectado play an educational role in creating greater student interest in learning new technologies. In the "Todo Chile Comunicado" project, they also performed training activities for the use of internet and computer. “Internet rural” regional projects managed by FITEL in Peru also include training programs aimed at providing basic knowledge to facilitate the use of the installed infrastructure. In Colombia, according to MinTIC, citizens and microenterprises do not see the usefulness of internet because of the lack of contents in native languages and local services or applications for citizens, or national microenterprises, as also the lack of appropriate technology. Free internet-based messaging mobile services, such as WhatsApp, are making SMS texting obsolete and raising concerns about it as a future revenue sources. Although something similar could happen with the introduction of digital voice (VoIP) services, the MNOs have reacted by offering competitive tariff plans including unlimited calls or a certain amount of minutes for a monthly rate.

12 www.grameentelecom.net.bd/

Page 49: TUCAN3G business model: case study and field verification

ICT-601102 STP Document number: D34 Title of deliverable: TUCAN3G business model: case study and field verification

TUCAN3G_D34EHASd.docx 49

5 SCALING UP THE SOLUTION When existing radio tower infrastructures are available (for site sharing) in the jungle areas, MNO can easily assume the mobile service deployment without external financial support, because they recover the low initial capital investment in the first year. Therefore, the easiest and faster way to scale up the TUCAN3G solution to other communities is a partnership with the owner of the existing rural network infrastructure.

5.1 Leveraging existing infrastructure

TUCAN3G proposes a partnership business model between a mobile services provider (MNOs) and a rural network operator (RNO) providing the backhauling service. Sharing available infrastructure is a win-win scenario for the MNO and the RNO. A backhaul service provider (RNO) cannot provide 3G mobile (public) services without a partnership with a licensed mobile operator (MNO). On the other hand, sharing an existing network infrastructure for backhauling generates enormous cost savings (CAPEX and OPEX) for the MNO. This partnership model makes possible (financially sustainable) to provide mobile services in rural communities with much fewer users than the traditional satellite approach. The local RNO can even operate (and maintain) the mobile access network (femtocells) for the MNO, including the provision of electric energy (solar power system). This is a major concern to ensure the smooth performance of mobile services that could otherwise be unsettled due to lack of reliable energy supply in isolated rural communities. The MNO will only provide a gateway access, so that the RNO can interconnect the rural mobile network to the mobile core network.

This partnership for sharing existing rural network infrastructure between the MNO and RNO implies ‘unbundling’ the traditional Telco business model. Today, this ‘unbundled’ business model is becoming the norm for mobile operators because infrastructure sharing generates many benefits [Deloitte15]. Mobile operators, such as Telefonica, are selling off passive infrastructure (radio towers) to infrastructure operators (TowerCos) and leasing back the use of them at lower total cost. This opens the door to a new business model, mobile infrastructure operators -TowerCo- and greater cost efficiency in network deployments. The logic behind this decision is a change in the competitive strategy, so far focused on network infrastructure and service management, to transform into a model of excellence in customer relations and offering a full range of integrated services, which can be developed by third parties. A typical TowerCo provides site sharing, which usually includes land (space on the ground to build the tower), masts and towers (steel), power supply, protection, sheltering and other passive equipment. The MNO installs the telecommunications equipment for both access and backhaul network using the passive infrastructure provided by the TowerCo. Noteworthy is the case of Peru, the new law that defines the role of RMIO, forces the MNO to negotiate the

Mobile Network Operator (MNO)

Rural Network Operator

Rural mobile coverage

$R (lease rate)

Figure 14: Partnership model with a Rural Network Operator

Page 50: TUCAN3G business model: case study and field verification

50

interconnection to its core network and allow the use of their licensed spectrum to provide mobile services (if they have not deployed their own infrastructure in a rural community). In Mexico, the single wholesale mobile network -Red Compartida- is another good example. As the mobile market grows, MNOs have increasingly outsourced the management of their network towers to a TowerCo in order to focus more heavily on their customers. TowerCo companies currently operate 18% of the mobile network towers in Africa, but are expected to operate up to 60% by 2020. Meanwhile, in India, tower companies already own up to 90% of the country’s towers. The market for tower constructions is currently valued at 20.3 billion dollars and is expected to grow as global tower installations are growing at a 4.1% 5-year CAGR. The tower company industry in Peru has significant room for growth. Peru currently has around 10.000 telecom towers, 15% of which belong to tower companies. By dividing the total number of Peruvian telecom towers by their population, we find Peru has a tower density per population (TDP) of .0265%. Peru’s TDP is lower than India, China, and Colombia’s TDPs, which are .06%, .05%, and .04%, respectively. In order to achieve 80% 3G coverage, Peru would have to construct 3.000 more towers. According to a forecast released by the Peruvian government, tower installations are expected to grow at a 9.86% CAGR over the next five years. The government forecasts Peru having 22.000 towers and a TDP of .064% by 2020 [Argandoña14]. According to data given by Agencia Nacional del Espectro ANE, (National Spectrum Agency), there are among 12.000 and 15.000 cell towers sites in Colombia. However, it will be required between 4.000 and 7.000 new cell in order to satisfy the growth of user demand. This will represent investments of more than $1 billion in a short time, and a wide spectrum of business opportunities in the country. Research from GSMA indicates [Kassam14] that future growth of mobile subscriptions will be concentrated in isolated rural areas where access to electricity is highly unreliable or not existent. If mobile operators extend their networks into remote locations to achieve universal coverage, the GSMA estimates that 160,000 additional towers will be needed by 2020 in these off-grid and bad-grid locations worldwide.

Table 24: Off-grid and bad-grid towers estimates by region (source GSMA, 2014)

Global estimates by region 2014 2020

Off-grid Bad-grid Total Off-grid Bad-grid Total South Asia 81,800 176,500 258,300 94,900 194,900 289,800 Sub-Saharan Africa 145,100 84,300 229,400 189,100 106,500 295,600 MENA 0 69,200 69,200 0 76,300 76,300 Latin America and Caribbean 58,400 265,600 324,000 62,500 288,400 350,900 East Asia and Pacific 34,800 105,400 140,200 43,300 125,000 168,300 TOTAL 320,100 701,000 1,021,000 389,800 791,100 1,180,900

Peru rural coverage gap

The map in Figure 15 shows that there are many populated areas (in brown colour) without any mobile infrastructure (coloured points represent cell tower sites). That is especially true in the Loreto department, which is accordingly the region in Peru with the lowest density in mobile subscriptions. The underserved rural areas are not attractive for mobile operators because they have low-income population, are inaccessible by wired terrestrial networks, lack reliable electricity supply and in the jungle the storms can cause damage to infrastructure. In summary, providing communications services in these regions imply higher costs and lower revenues.

Page 51: TUCAN3G business model: case study and field verification

ICT-601102 STP Document number: D34 Title of deliverable: TUCAN3G business model: case study and field verification

TUCAN3G_D34EHASd.docx 51

According to TowerXchange research13, the operators own the majority of cell towers in Peru, but there are also Torres Unidas and American Tower own more than 10% of the tower infrastructure in Peru. The new regulatory framework in Peru will facilitate the deployment of new telecoms infrastructure in rural areas. The estimated need of about 22,000 new towers within the next years is driving the tower market growth. Subsidies to operators that are expanding their networks in rural areas and coverage conditions imposed by the government in license renewals are facilitating that the underserved rural districts become “connected”.

Figure 15: Cell tower sites (coloured points) and population (brown area), Source MTC Peru Finally, there is other wireless network infrastructure (towers) deployed by PUCP and EHAS in isolated rural areas –about 144 radio towers- that can be shared to provide 3G mobile services as proposed by the TUCAN3G solution. As it is shown in Figure 16, this rural network infrastructure was deployed in areas where there is no other existing mobile infrastructure and was meant to meet basic needs in the poorest areas of Peru.

Figure 16: TowerCo cell sites and other rural network infrastructure

13 http://www.towerxchange.com/peruvian-tower-market-primed-for-significant-growth-in-next-three-years/

Page 52: TUCAN3G business model: case study and field verification

52

Colombia rural coverage gap

As seen in Figure 17, with the exception of Valle del Cauca, all departments in the Pacific region of Colombia have many areas where mobile coverage is almost non-existent. In the specific case of Cauca, there are coverage difficulties especially in municipalities located on the Pacific coast and the area named BotaCaucana, which is the most underserved zone in the Cauca department.

Cauca Valle del Cauca Chocó Nariño

Figure 17: Mobile coverage in Pacific region departments (source: OpenSignal, 2016)

The coverage maps (Figure 18) show that, in general terms, the departments of the Orinoquía region has deficient mobile coverage in the whole territory. The most complicated cases are in departments such as Arauca and Vichada, where there is less coverage area; the other departments show more coverage, at least in main areas.

Arauca Casanare Meta Vichada

Figure 18: Mobile coverage in Orinoquia region departments (source: OpenSignal, 2016) As shown in Figure 19, the six departments from Amazonia region have low coverage levels in the territory. The mobile phone coverage is possible in some main areas, but the rest of territory does not have reliable access to the mobile phone service.

Putumayo Caquetá Guaviare

Page 53: TUCAN3G business model: case study and field verification

ICT-601102 STP Document number: D34 Title of deliverable: TUCAN3G business model: case study and field verification

TUCAN3G_D34EHASd.docx 53

Guainía Vaupés Amazonas

Figure 19: Mobile coverage in Amazonia region departments (source: OpenSignal, 2016) The mobile service offered Claro and Movistar in Putumayo is not of practical use because of the low speed in data connections. Often, users prefer Wi-Fi connections, because getting WhatsApp or emails messages could take up to two hours.

5.2 Outsourcing network operation

Towards the sharing infrastructures objective, mobile operators are also striking network sharing deals with competitors or outsourcing network operations altogether to equipment manufacturers. When telecom equipment suppliers become a key partner for the mobile operator, they undertake main activities including network infrastructure operation, maintenance and (technical) services provisioning. By so doing, MNO’s costs are reduced (MNO, in turn, can focus on other activities regarding customer insight and branding). Equipment manufacturers such as Nokia or Ericsson run the networks at lower cost because they service several Telcos at a time and thus benefit from economies of scale. In a similar way, an RNO can outsource the network management to an equipment manufacturer. In this case, the role of RNO would be a partnership between an Equipment Manufacturer and a TowerCo (providing site sharing).

Mobile Network Operator (MNO)

Rural Network Operator

Site sharing (collocation)

$R (outsourcing)

Equipment Manufacturer

$R (lease rate)

Network managed service

Figure 20: Partnership model with an Equipment Manufacturer and a TowerCo

Page 54: TUCAN3G business model: case study and field verification

54

5.3 Building new infrastructure

In new deployments, especially in jungle areas, the high initial capital expenditure (CAPEX) may require a partnership with other public (or private) agents – i.e., Broadband Development National Agencies, Venture Capital Funds, Social Impact Investors, and International Development Agencies – that can subsidize the new deployments. The percentage of subsidies will depend highly on geographical conditions: a 60% of subsidies would be required to make the initiative sustainable in 5 years, in a 339 inhabitant’s community of a jungle area (where towers of 75m height are required). On the other hand, a 50% of subsidies would be required to make the initiative sustainable in a 100 inhabitant’s community in the mountains area (where towers of 15m are enough). NGOs can play an important role in this business approach, because obtaining a Social Return On Investment is a major concern for an increasing number of national and international agents when providing subsidies. Therefore, NGOs can apply for subsidies addressed to basic services infrastructures in underserved areas, and guarantee that the funds are employed to improve live conditions in these areas. Specifically, NGOs can use existing communication infrastructure to provide communication services specially designed to address local needs (agricultural, gender, medicine, education, etc.), providing an aggregated value (as happens with the pilot deployment at Napo River). Although this is the most expensive scenario (as no existing infrastructure is available), it can also be an opportunity to stimulate local entrepreneurship. Given that people living in isolated rural areas have a strong sense of community, the above-mentioned role of NGO could also be a Cooperative14 deploying community networks. These legal entities are autonomous associations founded by a set of stakeholders (in our case, the rural community people) who voluntarily cooperate for their mutual social, economic and cultural benefit.

Figure 21: Partnership model with an Impact Investing Agency to subsidize new deployments

14For coops in Latin America see http://www.aftic.gob.ar/institucional/nuevos-recursos-para-cooperativas-de-telecomunicaciones_n924 (in Spanish), http://www.srcoop.com.ar/

Page 55: TUCAN3G business model: case study and field verification

ICT-601102 STP Document number: D34 Title of deliverable: TUCAN3G business model: case study and field verification

TUCAN3G_D34EHASd.docx 55

6 CONCLUSIONS The TUCAN3G project aims to propose techno-economical solutions to break the digital gap and provide voice and data services (3G) in isolated areas of developing countries with less than 400 inhabitants. The technical solutions proposed to reduce costs of access and backhaul networks have proved its feasibility in a pilot deployed in two Peruvian Amazon areas, and this document has analysed the economical results obtained in the field trial. The first step was to perform a cost analysis to prove that TUCAN3G proposal has a lower Total Ownership Cost in 5 years than traditional solutions (based on satellite backhauling). Then we tried to characterize the demand of mobile services in rural areas of developing countries, and specifically, in jungle communities of Peru (where the pilot was deployed). In order to do so we conducted a market survey in four pilot communities, which served as baseline for impact evaluation and as a mean to estimate expected revenues. Despite the low number of users, which will grow steadily as it happens in other developing countries; initial data show that the ARPU ($ 11) in these locations could be even higher than the national average ($ 9). Being isolated, underserved rural communities, those who can afford to have a mobile phone know the real value of being connected both for personal and business relationships. Although the TUCAN3G pilot has been working only for 3 months (data of at least 6 months would be more conclusive),we have shown that, in most cases, the TUCAN3G solution can create value in less than 5 years. This period is quite reasonable taking into account that depreciation indexes in this sector are close to 5 years (communications technology becomes obsoleted quite fast). Moreover, the 10% of ROI and the positive NPV to be obtained in the first year (with subsidies for the cost of radio towers) are very promising. These results confirm the opportunity for scaling up the solution to other isolated communities, and allow us to go a step further and propose a business model for TUCAN3G solutions. However, providing communication services in isolated rural areas is a complex task that depends on a wide variety of factors such as regulation framework, geographical conditions, communication towers availability, funding options or target community size. Taking TUCAN3G pilot results as reference, this study has proposed three different business strategies to be selected depending on context conditions. In jungle communities with communication towers available, like in TUCAN3G scenarios, the MNO doesn´t need to establish partnerships with additional actors because the initial investment is low. The same happens in highland communities where the size of the tower (and the corresponding investment) is low (small towers up to 15m could cost $ 16,000 approximately) and the Net Present Value becomes positive in one or two years. When new towers are needed in jungle scenarios, deployments will require a high initial investment (CAPEX) to build new towers(up to 60,000 USD for radio towers of 90m would be required in jungle deployments).Therefore, MNO would probably require public incentives to provide cellular services in isolated communities of the jungle. These public incentives could have the form of subsidies, tax exemptions, license granting or regulation changes, and this documents provides some clues on how they could impact on operator’s business model. At this point, it is important to remember that governments are responsible for guaranteeing citizen’s rights, even (or specially) in isolated rural areas without infrastructures. Communication technologies can contribute to improve live conditions in rural communities by fostering economic development and by supporting basic services such as education, health or governability. Therefore, considering the Social Return On Investment of communication services should be a must when discussing the necessity of public support for this kind of initiatives. From the MNO perspective, stablishing multi-stakeholders partnerships would be an advisable strategywhen there are high risks of investment, because a clear advantage of multi-stakeholders

Page 56: TUCAN3G business model: case study and field verification

56

partnerships is that the risk is shared. MNO reduce their risk when working with rural networks providers, who have lower operating costs, and therefore, can be more cost-efficient when managing local networks. These private companies may also receive subsidies from public funds for development (national or international agencies) to reduce initial investment barriers. On the other hand, social civil organizations (NGOs or cooperatives) that have deployed telecom infrastructure in rural areas, such as EHAS-PUCP, can leverage their assets and obtain a periodic income to cover the operating costs. The conclusions here presented serve to demonstrate that it is technically and economically feasible to provide cellular services in rural communities of developing countries. Although the specific figures belong to a pilot in the Peruvian jungle, the same methodology can be applied in other countries. Moreover, telecommunication market follows quite similar trends in very different countries (in aspects such as cost structure, regulation or even demand). Therefore, these results should at least serve to foster research and investment in communications services for other isolated rural communities, which are looking forward to communicate with their family, friends, partnersor providers. In order to contribute to attend this social demand, TUCAN3G consortium will continue to support and extend the reference pilot in order to enlarge the dataset and provide clearer insights of how to provide 3G services in isolated rural communities of developing countries.