Mobile apps at the base of the pyramid ghana

Mobile at the Base of the Pyramid: Ghana

Growing Innovation

Mobile

9064_Ghana_Cover.indd 1

Š2014 infoDev / The World Bank | 1818 H Street, NW | Washington DC, 20433 Email: info@infoDev.org | Tel + 1 202 458 8831 | Twitter: @infoDev www.infodev.org

8/26/14 10:08 AM

Mobile at the Base of the Pyramid: Ghana

Mobile

9064_Ghana.indd i

8/26/14 10:03 AM

© 2014 International Bank for Reconstruction and Development / The World Bank Mailing Address: MSN I9-900 1818 H St. NW, Washington D.C., 20433 USA Telephone: (+1) 202-458-4070 Website: www.infoDev.org Email: info@infodev.org Twitter: @infoDev Facebook: /infoDevWBG Some rights reserved.

This work is a product of the staff of infoDev / World Bank. Note that the World Bank does not necessarily own each component of the content included in the work. The World Bank therefore does not warrant that the use of the content contained in the work will not infringe on the rights of third parties. The risk of claims resulting from such infringement rests solely with you. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of the donors of infoDev, The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved.

All photographs courtesy of The World Bank.

Rights and Permissions This work is available under the Creative Commons Attribution 3.0 Unported license (CC BY 3.0) http:// creativecommons.org/licenses/by/3.0. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Attribution—Please cite the work as follows: infoDev. 2014. Mobile at the Base of the Pyramid: Ghana.

Washington, DC: World Bank. License: Creative Commons Attribution CC BY 3.0 Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation.

All queries on rights and licenses should be addressed to infoDev, The World Bank, MSN: I9-900, 1818 H Street NW, Washington, DC 20433, USA; email: info@infodev.org

9064_Ghana.indd ii

8/26/14 10:03 AM

Acknowledgments

This study, “Mobile at the Base of the Pyramid,” was commissioned by infoDev, a global technology and innovation program at the World Bank. The study was carried out under the supervision of Maja Andjelkovic of infoDev and prepared by Research ICT Africa’s lead researchers, Steve Esselaar and Christoph Stork.

This report was made possible by the support of the Swedish International Development Cooperation Agency.

iii

9064_Ghana.indd iii

8/26/14 10:03 AM

Table of Contents

List of Acronyms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vi Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 BoP in Ghana . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Evolution of App Adoption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Smart and Feature Phone User Survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Business Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Basic Business Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Freemium Business Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Multisided Platform Business Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Revenue Sources—Paths to Monetization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 USSD/SMS-Based Content or Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 App Purchases, App Upgrades, and In-App Purchases . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 In-App Advertisements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Freemium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Subscriptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Other Revenue Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 Payment Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 Premium SMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Airtime . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Direct Carrier Billing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Mobile Money . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Vouchers and Scratch Cards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

iv

9064_Ghana.indd iv

8/26/14 10:03 AM

Distribution Channels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Operator App Stores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 OS App Stores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Third-Party App Stores and Websites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Conclusions and Recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Annex A. Mobile Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

v

9064_Ghana.indd v

8/26/14 10:03 AM

List of Acronyms

API app BoP CPC CPM DCB ICT IFC IP MEST MNO OS RIA RoP SMS USSD VAS VAT

application programming interface mobile application Base of the Pyramid cost per click cost per thousand impressions direct carrier billing information and communication technology International Finance Corporation Internet protocol Meltwater Entrepreneurial School of Technology mobile network operator operating system Research ICT Africa rest of the pyramid Short Message Service Unstructured Supplementary Service Data value added services value added tax

All dollar amounts are U.S. dollars unless otherwise indicated.

vi

9064_Ghana.indd vi

8/26/14 10:03 AM

9064_Ghana.indd vii

8/26/14 10:03 AM

Introduction

Apps may be used as stand-alone applications or to support or complement products or services.

A third usage is content delivery of some sort.

Products and services offered via mobile phones can further be differentiated as mobile services, mobile applications, and the mobile web. • Mobile services include premium Short Message Service (SMS), USSD,1 and voice services, which can be offered to any mobile user, including basic phone users. These services fall under the scope of the study, as they cater for 80 percent of base of the pyramid 1 USSD stands for Unstructured Supplementary Service Data. USSD is a protocol used by GSM cellular networks to communicate with the mobile network operator servers. The most common usage of USSD is for mobile money services. Unlike SMSes, USSD messages create a real-time connection to the server during a session that remains open, allowing a two-way exchange of a sequence of data.

(BoP) users, who are mainly basic phone users.

They may be integrated with mobile apps and mobile web services. • Mobile web services are services provided through mini browsers, such as Opera Mini. • Mobile applications are software packages that run on a feature and/or smartphone with an associated operating system (OS), such as Android, iOS, and so forth.

All three services belong to the same ecosystem and may be combined in a set of products. An app may use premium SMS as input from customers to deliver analytics to a business via a mobile application or mobile web. Mobile web per se would be out of the scope of this study, but it is included where used in the context of mobile apps or services. Most websites have a mobile web view, but including mobile web in the mobile app study would mean including most e-commerce sites.

2

9064_Ghana.indd 2

8/26/14 10:03 AM

9064_Ghana.indd 3

8/26/14 10:03 AM

BoP in Ghana refers to the infoDev definition of $2.5 per person per day.3 Individuals (adults of 15 years or more) were classified in terms of average household income, that is, household income divided by household size.

This section draws on a nationally representative household and individual ICT (information and communication technology) access and usage survey conducted by Research ICT Africa in 2012 in Ghana.2 For the purposes of this report, the BoP

Less than a quarter (22 percent) of the BoP has access to a bank account and about half own a mobile phone. Of those who own a mobile phone, more than 97 percent are prepaid. Of those with

2 Research ICT Africa, “Household and Small Business Survey” (2012), http://www.researchictafrica.net.

3 infoDev. “Mobiles at the Base of the Pyramid,” infoDev Project Concept Note, August 2011, http://www.infodev. org/infodev-files/resource/InfodevDocuments_1114.pdf.

TABLE 1: National Representative Survey Results for 2012 Ghana

BoP

%

Number

%

Number

Do you have a bank account (bank or post office)?

29.4%

4,365,279

21.8%

1,952,149

Do you own a mobile phone?

59.5%

8,824,392

51.2%

4,592,415

Prepaid

97.4%

8,593,187

97.1%

4,457,532

Is your mobile phone capable of browsing the Internet?

28.5%

2,511,999

22.7%

1,041,302

Facebook,Twitter, Mxit, or other social networking

11.3%

999,684

8.7%

397,443

Browsing the Internet

13.4%

1,181,060

10.9%

501,390

12.7%

1,878,455

10.3%

923,210

First used the Internet on a mobile phone

29.5%

1,324,196

30.5%

641,233

First used the Internet on a computer or laptop

70.5%

554,259

69.5%

281,977

Used the Internet on a mobile

61.2%

1,148,900

60.3%

556,355

Used the Internet at work

34.6%

500,228

23%

156,656

Used the Internet at a place of education

50.9%

735,917

62.1%

422,844

Used the Internet at Internet café

84.7%

1,225,160

93.8%

638,288

Signed up for any online social network (Facebook, etc.)?

80.6%

1,166,011

72%

489,954

Among 15+ mobile phone users

Do you ever use the Internet? (Gmail, Google, Facebook, Mxit, email)

Among 15+ Internet users

Source: Research ICT Africa.

4

9064_Ghana.indd 4

8/26/14 10:03 AM

a mobile phone, only 11 percent use it to browse the Internet. Internet use among the BoP is 10 percent, which includes non-mobile Internet use. The majority of Internet users started using it at an Internet café or a place of education.

What can be concluded from this for mobile application development?

• Twenty-three percent of BoP mobile users had a handset in 2012 that was capable of browsing the Internet and about 10 percent of BoP mobile phone owners used their mobile to actually browse the Internet. These figures may have increased since the survey was conducted at the beginning of 2012, because smartphone

and feature phone prices have dropped considerably. An entry-level smartphone is currently available for as little as $70.4 • Mobile money was at the early stages of adoption at the time of the 2012 survey—usage was approximately 1 percent. It is clear that about 78 percent of people at the BoP do not own a bank account and that SMS or airtimebased payments are the only feasible way for them to pay for apps or services at this stage.

4 Margaux Pelen, “What a $70 Smartphone Means for Mobile in Africa,” Medium (October 28, 2013), https:// medium.com/what-i-learned-today/99674f8d4f6f.

5

9064_Ghana.indd 5

8/26/14 10:03 AM

Evolution of App Adoption

Mobile application (app) adoption can be graded in three stages for people living at the BoP. • Stage 1—Basic phone. Basic phones are used and services can only be provided via SMS or USSD. About 80 percent of BoP mobile phone users in Ghana use a basic mobile phone that cannot be used to browse the Internet.

The distribution platform for basic phones is entirely in the control of the mobile network operators (MNOs). MNOs together with thirdparty facilitators take 85–95 percent of the after-tax SMS fee, leaving developers a mere 4–15 percent.

• Stage 2—Feature phone. The increasing prevalence of feature phones in stage 2 weakens the control of the MNOs because

it opens up the possibility of third-party app stores such as the Nokia Ovi Store, which is targeted at feature phone users and is able to negotiate better payment terms with operators by aggregating volumes via its app store.

Prices for feature phones and smartphones are declining and penetration within the BoP is increasing. Feature phones, for example, open up another potential revenue source: webbased services that are available through mini browsers, such as Opera Mini. Stage 2 has the advantage of the developer having more control over distribution and, therefore, revenue flows.

But control over revenue flows is premised on the availability of payment mechanisms other than premium SMS, such as mobile money, where developers no longer have to pay the high rates that MNOs charge for premium SMS payments. Mobile web typically charges via PayPal or credit cards. Advertisement for the mobile web is limited by small screen sizes. The various screen sizes also make it

FIGURE 1: Evolution of Mobile Adoption in Ghana Stages

Stage 1: Mobile voice & SMS

Stage 2: Mobile voice & data

Stage 3: Mobile computing

Revenue sources: Mobile operators

Airtime Share of premium SMS Share of mobile money fees

Airtime Share of premium SMS Share of mobile money fees

Airtime Share of premium SMS Share of mobile money fees

Revenue sources: Developer

Share of premium SMS

Web Share of premium SMS Mobile money

Share of app sales Mobile money revenue Web Share of premium SMS

Revenue sources: Third Party

Share of premium SMS Share of mobile money fees

Share of premium SMS Share of mobile money fees

Share of app sales Share of premium SMS Share of mobile money fees

Distribution channel

Mobile operators

Mobile operators Web

Mobile operators Web App stores

Technology

Basic mobile

Feature phone

Smart phone

Ghana:

approx. 77% of the BOP

approx. 20% of the BOP

approx. 0–5% of the BOP

6

9064_Ghana.indd 6

8/26/14 10:03 AM

difficult for developers to program apps for feature phones.

• Stage 3—Mobile computing. Smartphones open up a third revenue stream for developers: app sales, in-app sales, or in-app advertisement. Some MNOs operate app stores, but the most popular are operating systembased stores such as Google Play, iTunes, and BlackBerry App World. App developers can upload their apps to Google Play for free download, but they cannot directly arrange for paid downloads or in-app purchases. A way around this is to partner with developers in countries where Google Play has commercial partners, but there are none in Africa.5 Another option would be for an operator app store to collect the app price on behalf of Google in the Google store.

5 Google Play, “Supported Locations for Developer and Merchant Registration” (2013), https://support .google.com/payments/answer/1385282?hl=en&ref_ topic=1385074.

The ecosystem in stage 3 has widened to include additional players in the form of Google, Apple, BlackBerry, Facebook, and Nokia. However, app developers targeting the BoP need to consider that 80 percent of the BoP mobile phone owners are still in stage 1. This is changing rapidly, but the transition to stages 2 and 3 will take time. App stores usually take a 30 percent cut, leaving developers with 70 percent, a major improvement over premium SMS. Operator app stores that link to iTunes and Google Play may take a higher revenue share to cover the markup for their operating expenditure (OPEX) as well as the marketing of the app store. Smartphones in stage 3 allow developers the widest choice of revenue collection and control over distribution.

The smartphone penetration among mobile phone users of the BoP in Ghana is still very low— below 5 percent—but it is expected to increase because of Android-based smartphones costing under $70 coming on to the market.

7

9064_Ghana.indd 7

8/26/14 10:03 AM

Smart and Feature Phone User Survey

This section draws on a survey conducted of 800 mobile users by the research firm Jana6 in Ghana.

The survey was done entirely online. Users could access the survey page via mobile phone, desktop, or tablet. The description of survey respondents is displayed in table 2.

6 Jana is the world’s largest mobile research and rewards platform, covering 237 mobile operators. Registered panelists log on to mCent and see a list of surveys and offers available. If they participate, they earn airtime.

Respondents can access the survey via any means: computer, tablet, or mobile phone.

TABLE 2: Mobile User Survey among Smartphone and Feature Phone Users in Ghana Responses

800

Share female

Eighty-one percent of feature phone and smartphone users in Ghana preferred airtime deduction as a method of payment, pointing toward direct carrier billing (DCB). This highlights the low popularity of premium SMS, which only scored 4 percent. Because these users are early adopters of smartphones, it shows that premium SMS is most powerful when targeting basic and feature phone users. As people transition toward smartphones, new payment methods such as DCB and mobile money are vital to support innovation in the app ecosystem.

23.3%

Average age

24

Minimum age

15

Maximum age

61

Share rural responses

31.8%

Installed mobile app

93.8%

Previous app store use

The survey is not representative because participation was based on self-selection, that is, a user among the Jana user group can decide to participate or not, and there is no random selection. Only the first 800 responses per country were used. The majority of respondents were male and the average age was 23–24. About 93 percent of the respondents had already downloaded applications for their phones, so respondents can be classified as early adopters and indicative of a future trend. Sixty percent of respondents have used Google Play to download apps in Ghana, indicating that more than half the respondents are smartphone users.

Google Play

60.8%

Nokia Ovi

29.6%

FIGURE 2: Preferred Payment Method for App Downloads and In-App Payments by Feature Phone and Smartphone Users 81% Ghana

8% 8% 4%

MTN app store

20.8%

Samsung app store

15.5%

Other

Airtime deduction

Credit card

Mobile money

Premium SMS

n/a

Source: Survey commissioned by Research ICT Africa.

Note: n/a = not available.

8

9064_Ghana.indd 8

8/26/14 10:03 AM

9064_Ghana.indd 9

8/26/14 10:04 AM

Business Models FIGURE 3: Overlap of Choices

A business is required to realize four goals: solve a problem, make a profit, obtain customers, and keep customers. Applied to app development, a developer has to choose a problem to solve and think about how to make money solving it.

Revenues need to exceed costs and the first important consideration is revenue source.

Obtaining customers is about choosing the right distribution channel and using suitable payment facilities. Keeping customers is about continuously delivering value.

Distribution Channel

Payment Facilities

Revenue Sources

Decisions on distribution channel, revenue sources, and payment facilities are intertwined.

Most developers choose several combinations, either sequentially or simultaneously.

• Revenue sources include in-app advertisement, pay per download of content or apps, in-app purchases, and upgrade of a free app to a more feature-rich one in return for a fee and subscriptions.

• Distribution channels include choices to sell via an app store or directly to subscribers of an operator through value added services (VAS).

• Payment options include premium SMS, mobile money, airtime, credit or debit cards, and cash or checks.

The combination of these factors is reflected in various business models available for app developers. There are three categories of business models in order of complexity: basic, freemium, and multisided platforms. The basic business model includes three versions: app store, VAS provider, and content provider. The key feature of the freemium business model is that it is a blended model: free services are offered alongside paid premium services. The third business model comprises multisided platforms. The key feature of a multisided platform is that it facilitates transactions between two (or more) distinct yet interdependent customers.

Basic Business Models The three basic business models are displayed in figure 4: app store, VAS provider, and content provider.

App store The key feature of the app store model is that content or services are sold via an app store, such as Google Play or the Nokia Ovi Store. The developer gets a revenue share from the value of the app. For example, in the Google Play store, a developer would receive 70 percent and Google Play 30 percent. The same applies to inapp purchases.7 In this business model, the key questions that a developer should ask are the following: • How does the developer gain visibility in the app store?

• What is the payment mechanism? Google Play does not allow a business in any 7 Google Play, “In-App Billing Availability and Policies” (2013), https://support.google.com/googleplay/androiddeveloper/answer/1153481.

10

9064_Ghana.indd 10

8/26/14 10:04 AM

FIGURE 4: Basic Business Models

App purchase

Revenue Developer

App Store Share

Monthly Businesses

Customers in-App purchase

Bulk SMS VAS Provider

Customers

subscription

Content Provider

African country to register as a merchant, so developers have to go to other countries that are willing to act as intermediaries with Google.

• What is the revenue share with the app store?

The MTN App Store in Ghana has a revenue share split of 40/60, with 40 percent going to the developer. In comparison, the revenue share split for Google Play is 70/30, with 70 percent going to the developer.

• Are there any particular benefits of app store choices and/or requirements regarding exclusivity? For example, local app stores may promote apps and may require exclusivity in return.

An example of the app store business model is the iWarrior game from Leti Arts. iWarrior is a game

Premium SMS Bulk SMS

Customers

with African iconography. The purpose of the game is to protect a village from wild animals such as elephants, lions, and cheetahs. iWarrior is sold on iTunes for $0.99. Revenue share is 70/30, with 70 percent going to the developer.

VAS provider The key feature of the VAS provider model is that content and/or services are distributed via bulk SMS, which is paid for by businesses. In this model, the distribution channel is not an app store, but the MNO network. Key questions that a developer should ask: • Does the client provide the bulk SMS subscribers or are they accessed from a thirdparty database?

FIGURE 5: Business Model of iWarrior

Revenue share

OS store purchase App store

Customers

11

9064_Ghana.indd 11

8/26/14 10:04 AM

• How does the developer market to businesses?

Where does it find businesses that are interested in sending bulk SMSes to the developer’s customer base?

An example of the VAS provider business model is mNotify from Ghana. mNotifiy provides bulk SMS services to businesses, institutions, individuals, and clubs. Clients can send bulk SMSes through the mNotify web page. Subscription prices are linked to the quantity of SMSes; they vary from 5 cedis for 148 SMSes to 500 cedis for 17,858 SMSes (0.034 cedi to 0.028 cedi per SMS). mNotify also offers an application programming interface (API) to other app developers that allows them to include an SMSbased notification feature in their applications.

Clients pay for bulk services via formal banking facilities and mobile money. The purchase and activation of bulk SMSes is currently not automated. mNotify relies on Facebook for exposure. On average, it generates between $1,500 and $2,500 a month.

Nandimobile is an example of the content provider business model. Infoline is the main product of Nandimobile. It is a web and mobile application that enables organizations and businesses to communicate with their customers interactively.

Companies may send information to customers via bulk SMSes. Customers then respond to a short code number in order to request information or to provide feedback. Businesses may also advertise on billboards and in newspapers, for example, specifying a short code for potential customers to request more information. Infoline can be used for competitions and voting (TV game shows, for example). A premium SMS sent to any of the Infoline short codes from any network costs 0.1 cedi, of which 100 percent goes to mobile operators and third-party aggregators.

Nandimobile received $70,000 seed funding in 2010 and has been operating on revenues since then.

Distribution is mainly through the web and payments are collected manually from business customers.

This limits its scalability to services within Accra.

Content provider The key feature of the content provider model is that customers pay for content and/or services through premium SMS or interactive voice response (IVR). Premium SMSes are distributed via the MNO network. The content provider receives a small revenue share of the premium SMS value, but the main portion of the revenue from premium SMSes go to the MNO. Key questions that a developer should ask: • What is the revenue share split between the MNO and the developer?

• Will marketing be done by the MNO? If not, how will the developer ensure that customers are aware of their content or services?

Freemium Business Models Adding a layer of complexity are freemium business models. Freemium models are often used in conjunction with other business models, such as the content provider model. The key feature of the freemium model is that it is a blended model: there are free services alongside paid premium services. Generally, there is a large user base using the free app, which is crosssubsidized by a smaller user base using the paid version. (The ratio between free and paid users is often around 10 percent.)8 The paid subscribers 8 A. Osterwalder and Y. Pigneur, Business Model Generation (New York: John Wiley & Sons, 2010).

FIGURE 6: Business Model of mNotify

Monthly subscription Developer

m notify Wholesale

Bulk SMS

Customers

Bulk SMS Bulk SMS

other m-apps or resellers

12

9064_Ghana.indd 12

8/26/14 10:04 AM

FIGURE 7: Business Model of Nandimobile

Monthly subscription

Premium SMS Bulk SMS

Businesses

Customers

Advertisement

access both free and premium services (hence the “freemium” name). Key questions that a developer should ask: • What is the cost of providing content or services to free users?

• Do the fees charged to the paid user base cover both the free content as well as the cost of operation?

with real-time data. Farmers and traders can also match transactions and be notified by SMS. Aside from its main activity of providing market information to the agricultural sector, Esoko also offers a range of extension services through SMS. Basic services—such as market prices and bid and offer matching—are free, but additional services require either a subscription or are charged at premium SMS rates.

• How many paid subscribers does the application need to break even?

• How quickly or often do free subscribers convert to paid subscribers?

An example of the freemium business model is Esoko. Esoko is an agricultural platform that provides market information to farmers and traders. Esoko allows farmers to negotiate better prices with buyers by having real-time access to market prices in different locations and SMS alerts are automatically sent to farmers. At agricultural markets, prices are uploaded by either SMS, mobile application or website, providing farmers

Multisided Platform Business Models More complex combinations of the three dimensions (distribution channels, payment facilities, and revenue sources) can be found in multisided platform business models. Newspapers are a simple example of a multisided platform: newspapers can be sold below cost in order to attract more readers; in exchange, the newspaper can then charge more for advertisement space, effectively cross-subsidizing the newspaper

FIGURE 8: Business Model of Esoko Premium SMS to receive info

Monthly subscription

SMS alerts

Buyers Offer by free SMS

Farmers

Offers by free SMS Price information SMS, web or mobile app

Markets

13

9064_Ghana.indd 13

8/26/14 10:04 AM

FIGURE 9: Multisided Platforms for App Development

Businesses

Monthly subscription

Businesses, i.e. Advertisers

Advertisement

Multisided platform (e.g. mPedigree)

Free SMS/USSD

Multisided platform

Free App

readers. The key features of a multisided platform business model are as follows: • Network effect. The platform’s value is dependent on the number of users on each side. The larger the number of users on the one side, the more value that users on the opposite side see in the platform.

• Facilitator. A multisided platform facilitates transactions between the different customer bases.

• Distinct customers. The business model can differentiate between the different customer segments because distinct prices can be charged to each segment.

• Interdependent prices. Prices for the various sides are interdependent—that is, lowering the price for one side allows an increase of the price for the other side.

• Prices are set independently of cost. Allowing free download of an app is below the cost of developing the app. The zero price does not reflect the cost. The revenue to cover the cost is generated from the other side.

There are two versions of the multisided platform business model. In the first version, customer segment A gets the service for free (or heavily subsidized) while customer segment B pays for the service. In the second version, customer segment A pays to advertise to customers of segment B (usually consumers), thereby allowing the platform to subsidize the content to consumers.

Bulk SMS

Customers (of the app)

Customers (of the app)

Key questions that a developer should ask: • How can the developer attract sufficient A and B customers?

• Which side (or segment) is more pricesensitive?

• What is the optimal pricing for the interdependent markets? Free or low payment for one side and higher charges for the other, or more balanced payments?

Example 1: mPedigree One example of the multisided business model is mPedigree, which operates in Ghana and Nigeria. mPedigree is a verification service that allows users to send a free SMS to verify that a pharmaceutical drug is genuine. The cost of the SMS and the service mPedigree provides are paid for by pharmaceutical companies who have an interest in ensuring that the drugs people use are genuine and not counterfeit. In this variation of the multisided platform business model, the service to the consumer is free.

Example 2: mPawa mPawa is a job-matching service targeted at bluecollar workers in Africa. The majority of workers in Africa are blue collar and mPawa bridges the gap that exists in the current blue-collar recruitment space. mPawa is free for workers who register online or via SMS (mPawa therefore works on all phones); the workers then receive job notifications whenever prospective employers require their

FIGURE 10: Business Model of mPedigree

Pharmaceutical company

Subscription

Free SMS with drug code Patient Vertification SMS

14

9064_Ghana.indd 14

8/26/14 10:04 AM

FIGURE 11: Business Model of mPawa

10–50% of first month salary

Posting availability Blue-collar worker

Company Posting of jobs skill set. The main differentiating feature of mPawa is that, over time, a worker can build a strong résumé based on prior work and employer recommendations. The recommendation system reduces the time needed to find appropriately skilled workers. Workers are informed by SMS or they can log on to the mPawa website to see if they have been selected for a job. mPawa charges employers, not workers.

Employers are charged a percentage of the worker’s salary. Pricing depends on the length of the contract. A short-term contract (from one day to three months) costs 10 percent of the worker’s first month’s pay; a long-term contract (three months to one year), 25 percent of the worker’s first month’s pay; a permanent post (one year and above), 50 percent of the worker’s first month’s pay.

Example 3: Saya Saya, from Ghana, is a cross-platform chat application. Because it uses data, it is significantly cheaper than standard SMS. Saya works on both feature phones and smartphones and offers advanced functionality, such as Facebook chat for those with a Facebook account, group discussions, and the exchange of multimedia. Its most popular application is a location-based chat called street chat, where users can exchange messages based on their location. Saya was first offered only on feature phones, but it is now focusing on smartphones like Android and iOS because of their growing popularity. The value of Saya’s multisided platform is the size of its subscriber base and, therefore, its attraction to advertisers. Saya has no revenues, but it is testing an advertising model with AdsBrook, an app from the same incubator.

Job offers via SMS

Conclusions The best combination of these basic dimensions of payment facilities, revenue sources, and distribution channels will depend on the actual choices available in a country, the target audience, and the nature of product and services offered.

The majority of potential customers are still using basic phones. Feature phones and smartphones make up only 15–20 percent of the mobile subscriber base. Apps that are programmed for smartphones need to generate the required revenue from a smaller number of subscribers in order to sustain app development compared to an app developed for feature phones.

A multisided platform is the most promising business model for apps, particularly in countries such as Ghana, where payment facilities are limited or uneconomic. There are several examples of multisided platforms from the four countries, but most of the examples are from Nigeria and Ghana. mPawa is an example of a multisided platform that clearly solves a problem for two sets of customers: on one side it addresses employers, who struggle to find skilled and experienced workers; on the other side, it addresses the needs of blue-collar workers to find employment based on their previous work and skill set. The employers are relatively price-insensitive, so mPawa charges a fee based on the length of the employment contract. For workers, price sensitivity is high, so mPawa offers a free service.

FIGURE 12: Business Model of Saya Advertisement

Free app

Businesses

Data use Customers

Operator

Free chat

15

9064_Ghana.indd 15

8/26/14 10:04 AM

Revenue Sources—Paths to Monetization TABLE 3: Types of Revenue Sources

The decision concerning the generation of revenues is at the heart of any business model.

App developers have to make a choice about how to make money from their products and services, using fees and subscriptions to generate revenues, such as in-app advertisements, pay per downloads of content or app, in-app purchases, upgrades of a free app to a more feature-rich one, and so forth.

Globally, advertising and app purchases are the dominant mechanism for revenue generation (see figure 13). The trend for 2013 also shows a shift toward advertisements, in-app purchases, and the freemium models.9

In-app advertisements

Advertisers pay app developers to place adverts within the app (e.g., at the bottom of the screen).

Pay per downloads

The most common model, where subscribers pay a fee to download the app.

In-app purchases

Downloading the app is often free, but the user has to pay to use the app.

Freemium

Basic services are free. Advanced services (i.e., additional features) are charged.

Subscriptions

Frequent, regular payments to the app developer in order to continue using the app.

Developers in Ghana have, in principle, the same choices available as anywhere else, but they are

FIGURE 13: Share Revenue Sources Used among Global Mobile Developers (n = 2,167)9 38% Advertising

33%

32% Pay per download

34%

26% In-app purchases

19%

25%

Freemium

18% 12% 12%

Subscriptions 2013

2012

9 VisionMobile, 2013.

16

9064_Ghana.indd 16

8/26/14 10:04 AM

constrained by various market factors that limit their choices and their ability to raise revenues when products and services are targeting local (that is, Ghanaian) audiences.

• Smartphone penetration is still quite low in Ghana.

• MNOs own the distribution channels—via their SIM cards—to subscribers and take revenue shares of between 60 and 70 percent for content delivery.

• Local app stores are in the development stage, but they are either not yet launched or only a few months old.

The various revenue sources are discussed in the sections below.

USSD/SMS-Based Content or Services VAS services have been offered for many years in terms of ringtones, wallpapers, and daily horoscopes. Any developer considering providing services or content via this channel needs to be aware that the subscribers, the network, and the payment facilities are all in the hands of the Telcos. The remaining revenue share for developers is relatively small. At the same time, this is the most effective channel to reach large numbers of the BoP. Ideally, apps using this channel should ensure that premium SMS is not the only revenue source and that it is complemented by others.

App Purchases, App Upgrades, and In-App Purchases When programming for a global audience and launching products through Google Play, Facebook, iTunes, and so forth, the mechanics are in principle the same, independent of the location of the developer, but in practice there are limitations. By far the most attractive platform for

African app developers is the Google Play store: its revenue split is 70/30 in favor of developers and the procedures for registering an app are relatively straightforward and transparent. In contrast, the API for the MTN Ghana app store (the only local app store at present in Ghana) is only available upon request. Just getting access to the API of a local app store (for example, MTN) is an achievement aside from the conditions that are sometimes attached, such as exclusivity in the operator app store. Google Play’s API is available on the web, easily accessible, and easily understood with few conditions.

Though the Google Play store is popular, the list of countries in which developers may register as merchants does not include a single African country.10 This means that developers are either restricted to free apps on Google Play or have to work through partners registered as merchants on another continent. Another option is to sell through an operator app store such as MTN Ghana. Multinational operators may have an agreement with Google to act as a merchant for all the countries they operate in. Payment is then made from the user to the operator and from the operator to the developer. The revenue share is thus likely to be lower compared to directly selling through the Google Play store.

In addition to selling an application, revenues can be generated by selling upgrades with more features or allowing in-app purchases.

In-App Advertisements In-app advertisements represent a promising source of revenue for app developers in Africa.

The advertising model generally requires an intermediary to match advertisers with app developers or publishers. The attractiveness

10 Google Play, “Supported Locations for Developer and Merchant Registration” (2013), https://support .google.com/googleplay/android-developer/ answer/150324?hl=en&ref_topic=15867.

17

9064_Ghana.indd 17

8/26/14 10:04 AM

FIGURE 14: Business Model of AdsBrook

$2.5–$5 per

$0.95–$1.20 per

1,000 impressions

1,000 impressions

Company

of a publisher to an advertiser is correlated to the size of its subscriber base. This is the key characteristic of a multisided platform.

AdsBrook is such an intermediary, and it offers targeted advertisements within apps and online.

AdsBrook is specifically focused on providing local consumers with relevant ad content. The application allows developers to serve ads across different mobile platforms, namely Android, iOS, J2ME, and BlackBerry. Regional competitors include Twinpine Networks from Nigeria, and international competitors include Blueview .com and m-mobi. AdsBrook has chosen a cost per thousand impressions (CPM) model where an advertiser pays between $2.5 and $5 for 1,000 impressions, of which $0.95–$1.20 is paid to the app developer. The advantage of this model is that the app developer (for example, Saya) is guaranteed a relatively stable income, assuming that Saya can predict its traffic flow. The alternative model, adopted by Twinpine Networks, is cost per click (CPC), where an advertiser only pays when a user clicks on the advertisement.

Freemium The freemium model reduces risk by allowing consumers to test out the company’s product or service and establish whether they would purchase more features. The freemium model provides a basic product/service for free but charges a premium for value-added features. This is a successful model for the BoP market because it provides a risk-free trial and the low-income consumer can clearly understand the benefits of the additional services that the paid product would provide. This model has been most commonly used by agricultural information services, pioneered by apps such as Esoko.

Blue-collar worker

Subscriptions Monthly subscriptions are an attractive revenue source because of their stable cash flow.

Companies may pay for a range of services, such as delivering bulk SMSes to their customers, allowing customers to request information, informing farmers about market prices, and so forth.

Other Revenue Sources Revenues may be generated in a number of other ways. Apps may be commissioned by public institutions, such as schools and clinics or companies. Contract work can be used to subsidize the business in order to develop other applications (Leti Arts, for example, takes that approach). Apps may be designed to support business processes and to manage information and/or to receive and make payment. The revenue is then not generated by the app but by the supported business process.

Several apps that were interviewed for this research either required contract work to ensure their continued survival or used contract work as a mechanism to earn sufficient funds to be able to develop apps. For example, to support app development, Leti Arts does contract work, using African traditions and themes, to build employee orientation games, and a game for the United Nations explaining developmental milestones for new mothers.

Conclusions Internationally, there is a shift toward advertisements, in-app purchases, and freemium models as sources of revenue. In-app purchases

18

9064_Ghana.indd 18

8/26/14 10:04 AM

and advertising revenue are two of the most promising revenue sources in Ghana. While mobile advertising is nascent, it represents an exciting potential market as app developers build large subscriber bases.

The path to monetization in Ghana is extremely challenging. Several app developers have to

supplement their income through contract work in order to cross-subsidize app development. There are several revenue sources potentially available, but few are practically possible because of the lack of payment facilities. The challenge remains the ability to match revenue sources with available payment facilities.

19

9064_Ghana.indd 19

8/26/14 10:04 AM

Payment Options is neither bank- nor operator-based comprises scratch cards and vouchers.

The main payment options for app developers to collect revenues can be grouped into operator based, mobile money, bank based, and others.

Operator-based payment facilities include premium SMS and airtime transfer. Mobile money has been placed in its own category because it can be operator-based or bank-based, depending on the regulatory framework. Bank-based payments are electronic transfers or payment by check or with debit and credit cards. A fourth category that

The cost of collecting revenues is different for each of these payment facilities and so is their availability. For example, app developers receive between 5 percent and 15 percent of revenues using premium SMS and there is a potential target market of 8.8 million mobile subscribers (aged 15+) in Ghana. They would, however, receive 95 percent of revenue or more if they could use mobile money, but then their potential customer base would be limited to about 400,000 active users.

FIGURE 15: Payments Options for App Developers in Ghana

Payment Options

Ghana

Operator Based:

Operator Based:

Premium SMS

Developer revenue share: 5–15% Customer base 15 : 8.8 million

Airtime transfer

Developer revenue share: 85–90% Customer base 15 : 8.8 million

Mobile Money:

Mobile Money:

Mobile money

Developer revenue share: 95% Customer base: 400,000

Bank Based:

Bank Based:

EFT/Check

Developer revenue share: 95% Customer base 15 : 4.3 million

Credit/Debit card

Developer revenue share: > 95% Customer base: unknown

Other:

Other:

Vouchers/Scratch cards

Developer revenue share: 95% Customer base 15 : 8.8 million

20

9064_Ghana.indd 20

8/26/14 10:04 AM

The upper limit, in terms of number of potential customers, that an app developer can target is 8.8 million, which is the number of Ghanaians 15 years or older who owned a mobile phone in 2012. This figure will be higher in 2013 and continue to steadily increase. Using formal banking facilities, apart from being less convenient compared to just sending a SMS, halves the customer base, because only 4.3 million people have some banking facilities. The number of Ghanaians with a credit card or debit card could not be reliably established, but it is significantly less than those with bank accounts. Vouchers and scratch cards potentially target the entire mobile user base, but the distribution channel would need to be built up from scratch and the actual target market would be local, unless one wants to distribute the vouchers along the airtime channels of operators, which would entail a significantly lower revenue share.

Premium SMS Premium SMSes are ordinary SMSes, but instead of content they entail a payment instruction. The airtime of the subscriber sending a premium SMS is reduced by a specified value, usually higher than the cost of a standard SMS. Because these messages cost a premium fee, they are typically designated a special number (a “short code,� which is usually four to eight digits in length).

In some countries, usage of premium SMS can be restricted (and is, therefore, controversial). For example, in Zambia, the regulator charges $3,000 for use of a designated short code in an attempt to control content delivery. In Mozambique, short codes are allocated largely to MNOs, making it difficult for app developers to get access.

Premium SMS is the most attractive payment facility, because the target market consists of all 8.8 million mobile users. However, the economics of premium SMS mean that it is very difficult to

make money. In Ghana, two of the major SMS aggregators are SMSGH and Rancard. In tables 4 and 5, the economics of premium SMSes are illustrated for each business model.

SMSGH takes 25 percent of the after-tax revenue from premium SMSes, with the operators taking between 60 and 70 percent, based on a premium SMS cost of 5 pesewas. On the MTN or Tigo networks, an app developer would need to sell 2.4 million premium SMSes to have revenues of 10,000 Ghanaian cedis, or $5,000 per month. The total potential market is 8.8 million subscribers and so over a quarter of the Ghanaian mobile owners would need to subscribe, a nearly impossible target. With higher nomination premium SMS, the revenue share of SMSGH is lower and the mechanics start to look more positive for the developer. However, higher premium SMS prices also mean lower participation by subscribers.

Rancard’s economic model is slightly different and generally more lucrative for the app developer.

Rancard takes 60 percent of what is left after taxes and after operators have taken their share.

Nevertheless, revenues of 10,000 cedis, or $5,000, still require 1.7 million premium SMSes sold on the MTN or Tigo networks.

The business models outlined in tables 4 and 5 for Rancard and SMSGH are negotiable. Depending on the volume of premium SMSes, more favorable conditions may be accepted. However, the mobile operators in Ghana seem unwilling to negotiate.

That may change with revenue streams shifting from premium SMS to app stores.

The premium SMS economic model explains why premium SMS generally remains the preserve of mobile operators, with few app developers using premium SMS as a payment facility. While the number of subscribers is attractive, the requirement for massive volumes severely limits the choice of potential business models and the revenue share model has a chilling effect on app innovation.

21

9064_Ghana.indd 21

8/26/14 10:04 AM

TABLE 4: Premium SMS in Cedi, Based on SMSGH Business Model Premium SMS 50 Pesewas

Premium SMS 5 Pesewas

Vodaphone

MTN or Tigo

Artel

Vodaphone

MTN or Tigo

Artel

Premium SMS

0.05

0.05

0.05

0.5

0.5

0.5

VAT

15%

15%

15%

15%

15%

15%

6.09%

6.09%

6.09%

6.09%

6.09%

6.09%

21.09%

21.09%

21.09%

21.09%

21.09%

21.09%

0.041

0.04

0.04

0.41

0.41

0.41

Operator share

60%

65%

70%

60%

65%

70%

SMSGH share

25%

25%

25%

7.5%

7.5%

7.5%

Developer share

15%

10%

5%

32.5%

27.5%

22.5%

Share of nominal SMS value

12.4%

8.3%

4.1%

26.8%

22.7%

18.6%

Revenue for app developer

0.0062

0.0041

0.0021

0.1342

0.1136

0.0929

1,614,533

2,421,800

4,843,600

74,517

88,065

107,636

Communication tax Total tax After tax revenue

Premium SMS received to raise 10,000 cedis

Source: Author interview and calculations.

Airtime Airtime is, theoretically, an attractive alternative payment facility to premium SMS. It has the same appealing features: a potential subscriber base of 8.8 million and a payment method with which all subscribers are familiar. However, there are two challenges in using airtime: first, it can’t easily be integrated into an app; it requires a separate transaction, whereby the buyer has to send airtime to the seller and assume that the seller will honor the transaction. Second, mobile operators oppose the use of airtime as a payment mechanism11 and no developers have integrated airtime as a payment facility in Ghana (although there are some examples of using airtime as a payment facility in Nigeria). Also, airtime as a payment facility has the

11 One of the reasons mobile operators oppose using airtime as a payment method is that Central Banks tend to discourage it, because it can potentially become an alternative currency and it can hide money laundering.

challenge of cashing out: the accumulated airtime has to be resold in order to get cash.

Direct Carrier Billing Direct carrier billing (DCB) is when the consumer’s mobile airtime is directly deducted when purchasing an app. In comparison to premium SMS, DCB offers greater flexibility: the payment can be deducted in-app rather than via sending or receiving an SMS. DCB is integrated directly into the mobile network operator’s billing platform.

DCB is not available in Ghana yet. In terms of revenue share, DCB is far more attractive than premium SMS because operators typically receive between 10 and 20 percent of revenue rather than the 60 to 70 percent typical of premium SMS. Stakeholder organizations, such as the Wireless Access Service Providers’ Association of Ghana

22

9064_Ghana.indd 22

8/26/14 10:04 AM

TABLE 5: Premium SMS in Cedi, Based on Rancard Business Model Premium SMS 50 Pesewas

Premium SMS 5 Pesewas

Vodaphone

MTN or Tigo

Artel

Vodaphone

MTN or Tigo

Artel

Premium SMS nominal values

0.05

0.05

0.05

0.5

0.5

0.5

VAT

15%

15%

15%

15%

15%

15%

6.09%

6.09%

6.09%

6.09%

6.09%

6.09%

21.09%

21.09%

21.09%

21.09%

21.09%

21.09%

0.041

0.04

0.04

0.41

0.41

0.41

60%

65%

70%

60%

65%

70%

Funds remaining

0.0165

0.0145

0.0124

0.1652

0.1445

0.1239

Rancard share 60% of remaining

0.0099

0.0087

0.0074

0.0991

0.0867

0.0743

Revenue for app developer

0.0066

0.0058

0.0050

0.0661

0.0578

0.0495

Share of nominal SMS value for developer

13.2%

11.6%

9.9%

13.2%

11.6%

9.9%

Share of nominal SMS value for Rancard

19.8%

17.3%

14.9%

19.8%

17.3%

14.9%

1,513,625

1,729,857

2,018,167

151,363

172,986

201,817

Communication tax Total tax After tax revenue Operator share

Premium SMS received to raise 10,000 cedis

Source: Author interview and calculations.

(WASPAG), can play a role in lobbying MNOs to implement DCB.12

Mobile Money After premium SMS and airtime, mobile money is the most attractive payment facility. However, the number of active mobile money users (as opposed to mobile money subscribers) is low, at around 400,000 out of a total population of 25.37 million, 12 WASPAG was established in 2012 by leading companies within the mobile VAS and applications market. The purpose of the organization is threefold: to ensure industry stakeholders earn a fair return on their investments; to ensure that members work for the benefit of the association; and to provide a platform for addressing the concerns and complaints of the public (WASPAG, accessed December 2013, http://www .waspag.com).

representing a penetration rate of 1.5 percent.

Also, none of the mobile money initiatives have been technically structured to appeal to app developers: APIs are not publicly available.

A developer can get access to an API only via negotiation with each operator and usually only if the operator has assessed the business model of the proposed application.

Aside from the technical obstacles, two of the three mobile money operators in Ghana are not set up to process micro-payments effectively: both MTN Mobile Money and Airtel Money have a minimum charge of 1 cedi or $0.5, representing a transaction fee of 100 percent for 1 cedi payments.

Tigo Cash is a substantially cheaper service, with transaction fees of 3.5 percent on a transfer of 1 cedi, and can therefore handle micro-payments.

23

9064_Ghana.indd 23

8/26/14 10:04 AM

TABLE 6: Mobile Money Charges: MTN and Airtel MTN Mobile Money: Cost to Sender

Airtel Money: Cost to Sender

Transfer amount in cedis

1

25

50

75

100

1

25

50

75

100

Cash in

0

0

0

0

0

0

0

0

0

0

Transfer fee

0.5

0.5

0.5

0.75

1

0.5

0.5

0.5

0.75

1

Cash out

0.5

0.5

0.5

0.75

1

0.5

0.5

0.5

0.75

1

1.00

1.0

1.0

1.5

2.0

1.0

1.0

1.0

1.5

2.0

100%

4%

2%

2%

2%

100%

4%

2%

2%

2%

Total cost cash to cash Total cost cash to cash in %

Source: Author interview and calculations.

As mobile money penetration increases, Tigo Cash may represent a significant opportunity for app developers to integrate mobile money as a payment facility. To be appealing to an app developer, all mobile money solutions should ideally be integrated into the app.

Vouchers and Scratch Cards Vouchers and scratch cards are ideal for local prepaid services, but no app identified in Ghana uses vouchers. They are, however, used at Internet cafĂŠs. While not necessarily suitable for app stores, vouchers could be sold to activate

TABLE 7: Tigo Cash: Cost to Sender Registered User Transfer amount in cedis

Other

0.5

1

25

50

75

100

1,000

1

25

50

75

100

1,000

0

0

0

0

0

0

0

0

0

0

0

0

0

Transfer fee

1%

1%

1%

1%

1%

1%

1

3.5%

3.5%

3.5%

3.5%

3.5%

3.5

Cash out

2.5%

2.5%

2.5%

2.5%

2.5%

2.5%

2.5

0

0

0

0

0

0

Total cost cash to cash

0.02

0.04

0.88

1.75

2.63

3.50

3.50

0.04

0.88

1.75

2.63

3.50

3.50

Total cost cash to cash in %

3.5%

3.5%

3.5%

3.5%

3.5%

3.5%

0.4%

3.5%

3.5%

3.5%

3.5%

3.5%

0.4%

Cash in

Source: Tigo website.

24

9064_Ghana.indd 24

8/26/14 10:04 AM

subscriptions to mobile services that target mobile users in particular locations, such as farmers requiring specific market prices.

Conclusions Of the four countries in this study, Ghana has the worst environment for payment facilities: premium SMS rates are exorbitant. While revenue share splits can be negotiated, few app developers have been able to get better terms. Airtel, for example, refuses to negotiate premium SMS rates. MNO stores have a unique advantage in terms of having access to subscribers, though this advantage will

be eroded as smartphone penetration increases.

As the feature phone and smartphone survey shows, there is enormous appetite for direct carrier billing, and organizations like WASPAG can play a crucial role in advocating the introduction of DCB. Mobile money as a payment facility is still developing, and penetration will be slow while service providers build agent networks. Of the three major mobile money service providers, only Tigo Cash supports micro-payments.

For a developer wanting to use mobile money as a payment facility, Tigo Cash is the only viable option.

25

9064_Ghana.indd 25

8/26/14 10:04 AM

Distribution Channels

The availability of a mobile application is dependent upon its distribution network.

Distribution can take place via an app store, a mobile operator (through USSD services, for example), or a local distribution network. Using USSD or SMS-based services for content delivery has the advantage of the distribution channel already being in place through mobile operators.

This is the primary justification for operators taking their high revenue share. In addition, operators are actively marketing the VAS content on their platforms.

Apps developers also have the choice of operator app stores, operating system app stores, independent app stores, or local app stores.

Globally, the most common platforms are Android and iOS (see table 8). BlackBerry is dropping behind Windows Phone as its user base declines.

Operator App Stores In Ghana, only MTN has launched an app store, and it has been operational for only a few months.

TABLE 8: Primary Platform Usage among Mobile Developers Platform

% of Developers

Android

34.4%

iOS

32.7%

HTML5 mobile

17.3%

Windows Phone

4.5%

BlackBerry

4.2%

Other Source: VisionMobile, 2013.

7%

MTN is attempting to increase the range and relevance of apps by hosting the App Development Competition. MTN splits any revenue 40/60, with MTN receiving the larger share. In comparison, the Google Play store splits revenues 70/30, with the app developer receiving the larger share. MTN’s justification for the larger revenue share is that it promotes local apps via its network, something that Google Play is unable to do. However, there are hardly any apps that have a payment facility, so revenue share is a moot point. The 40/60 split is likely to be motivated by the high revenue shares that operators take for premium SMSes and ignores the different mechanics that characterize the app ecosystem, where the value chain is no longer fully controlled by MNOs. MTN will need to adjust its revenue share to global standards in the medium term or lose this business altogether.

None of the other mobile operators in Ghana have launched an app store yet.

OS App Stores For the smartphone market, international stores such as Google Play and BlackBerry World constitute the primary distribution platform.

Uploading on Google Play is free and revenues are split 70/30. There is no payment facility available on the African continent for Google Play, so apps that are targeting Africans are generally free, unless developers have set up agreements to work through partners registered as merchants with Google Play on another continent.

BlackBerry World (formerly BlackBerry App World) is the app store for BlackBerry devices.

BlackBerry World has 120,000 apps, and payment can be made by credit card and PayPal (some countries can pay by direct carrier billing but this is not available in Ghana). BlackBerry World is the third largest app platform in the world, after iOS and Google Play.13 Due to its efficient compression 13 VisionMobile, Developer Economics 2013. Developer Tools: The Foundations of the App Economy, http://www .visionmobile.com/product/developer-economics-2013the-tools-report/.

26

9064_Ghana.indd 26

8/26/14 10:04 AM

technology—resulting in lower bandwidth usage—BlackBerry is one of the most popular smartphones in Africa and is still seen as a viable development platform, despite its declining user base in Europe and North America.

Third-Party App Stores and Websites GetJar is the major third-party OS store available to the Ghanaian market. GetJar is a third party because it is independent of handset operators (such as Nokia) and the OS stores (such as iTunes, Google Play, and BlackBerry World). Uploading apps onto GetJar is free and there is no revenue share requirement. From a revenue share perspective, this is an attractive app store, but its awareness among BoP consumers is reportedly low.14 There are also websites that act as content aggregation platforms to which users can log on to get apps, music, and videos. Examples include www.waptrick.com and www.sefan.ru.

An alternative distribution method is an app developer’s own app store. This has the advantage of selling new apps to consumers who are already

14 Comments from Zubair Alhaji Abubakar, co-founder of Pledge51.

aware of the publisher, and who both like and trust the content. In order to host the various games and comics that it develops, both in-house and for other companies, Leti Arts launched Leti Centre in October 2013 (available for both feature and Android phones). This is essentially an app store that enables Leti Arts to provide a range of different games and comics within one application and to build its subscriber base outside the realm of telecom operators. The publisher-created app store is a relatively new strategy, and publishers/ app developers will need to invest in making app stores an actual resource, ensuring that apps receive relevant recommendations and reviews as well as timely updates.

Conclusions There is one operator app store in Ghana— MTN—and it was only recently launched. While it builds subscriber numbers and brand loyalty, app developers may be able to negotiate more beneficial revenue splits. Competitions, such as the App Developer Competition from MTN, offer one tool to increase the supply of apps, but one-off competitions are unlikely to be the basis for a dramatic change in the availability of apps and the number of downloads. The challenge is that too few apps are using the available payment

FIGURE 16: Waptrick Website

27

9064_Ghana.indd 27

8/26/14 10:04 AM

facilities, such as premium SMS, because of the perceived inequity in revenue share. The universal complaint among app developers was that revenue share was unfair and that payment, when it did take place, was often delayed for many months.

Operator app stores suffer from the archetypal causality dilemma: if app developers have a viable business model, then MNOs will offer good terms; if MNOs offer good terms, then app developers will develop apps. MNOs are preoccupied with the app business model and require the app developer to submit a business plan to even get the API. App

developers distrust MNOs and feel that they are not being treated fairly. The net result is inaction.

But MNOs are prepared to negotiate and there is a way that all parties can profit—it just requires communication and commitment from both sides. Google Play is currently only an option for free apps, where the revenue is made through advertisement or through other means. With the lack of local app stores, there is clearly room for innovation, but it is too early to tell if these alternative app stores will be viable and popular distribution methods.

28

9064_Ghana.indd 28

8/26/14 10:04 AM

9064_Ghana.indd 29

8/26/14 10:04 AM

Conclusions and Recommendations

The app environment in Ghana is quite constrained. While innovation is taking place both at the level of revenue models and distribution channels, the primary obstacle is the lack of viable payment facilities. The easiest payment option is premium SMS because it is available on all types of phones, is easily understood by the consumer, and is easy to implement. Mobile operators, however, have a lock on premium SMS and it suffers from poor economics from the point of view of the app developer. The poor revenue share—app developers receive between 5 and 15 percent— means that developers actively avoid premium SMS. The net effect is to discourage innovation where premium SMS is the only payment facility.

The poor economics of premium SMS is partly a result of the lack of alternatives. Operators know that there is little option but to use premium SMS. Airtime is a viable alternative, but it is generally discouraged by both operators and regulators.15 Mobile money penetration is very small and the market is fragmented.

The solution is to creatively investigate ways to bypass the payment facility problem and to find business models that can rely upon the formal banking sector to collect payments. In our exploration of the market in Ghana, the most successful business model has been a multisided platform. The defining characteristic of a multisided platform model is that it brings together two (or more) distinct but interdependent groups of customers. The platform is only valuable if the other group of customers is also present, in which case transactions between different groups of customers are facilitated. There are several successful apps in this case study that use a multisided platform.

This business model has the advantage of addressing a specific gap in the market for two sets of customers. As a result, awareness is easier and the distribution channel can be independent of the local app stores, whose revenue share is considerably higher than that of international app stores such as Google Play and GetJar. Revenue share for GetJar is free, but awareness is significantly lower than Google Play.

Without a micro-payment processing platform, apps within Ghana will continue to struggle.

While the multisided platform business model is a viable and successful approach, it generally relies upon businesses wanting to communicate information (often via advertising) to consumers and so dispenses with the business model where consumers are prepared to pay for the goods or services that they consume.

Recommendations for app developers include the following: • For those apps targeting the mass market, premium SMS is still the only viable option.

However, it is expensive. If the volume is high, some operators are prepared to offer better terms than the common 70/30 split.

Recommendation: developers with significant subscriber numbers should negotiate directly with operators to achieve better revenue share terms.

• For app developers that do not have a large subscriber base, aggregating with other developers or with third parties is the best way to enhance their negotiating power with MNOs. Incubators or hubs can also play a role in aggregating demand for premium SMS and further reduce the power of MNOs. At present, there is a high level of distrust between MNOs and app developers. Recommendation: create better communication channels (through incubators, third-party aggregators, and stakeholder associations) in order to reduce the levels of distrust.

15 Refer to footnote 11.

30

9064_Ghana.indd 30

8/26/14 10:04 AM

• In the current environment with limited payment options, multisided platforms provide the most exciting business model. There are several examples in Zambia and Nigeria that could be imported into Ghana as potential business opportunities. Recommendation: use multisided platform business models. • Advertising is a potentially profitable avenue.

Building a subscriber base by providing a free app on Google Play, for example, is one approach. Once a critical mass has been reached, businesses or third-party advertisers may be approached for ad placement. As apps collect data about their subscribers and the

retail sector continues to expand, there is a potential match between advertisers (such as retailers) and app developers. Although mobile advertising in Ghana is growing, this model faces a significant challenge: nearly all app developers have reported insignificant revenues coming from advertising. But advertising platforms such as AdsBrook are indicative of the potential of the market, and more apps should be investigating how to take advantage of the growth in mobile advertising.

Recommendation: investigate advertising as a potential revenue source.

31

9064_Ghana.indd 31

8/26/14 10:04 AM

References

Airtel Ghana website, accessed November 2013, http://africa.airtel.com/wps/wcm/connect/africarevamp/ Ghana/.

Google. 2013. “In-App Billing Availability and Policies.” https://support.google.com/googleplay/androiddeveloper/answer/1153481.

Google Play. 2013. “Supported Locations for Developer and Merchant Registration.” https://support.google .com/googleplay/android-developer/answer/150324?hl=en&ref_topic=15867.

infoDev. 2011. “Mobiles at the Base of the Pyramid,” infoDev Project Concept Note, World Bank, Washington, DC. http://www.infodev.org/infodev-files/resource/InfodevDocuments_1114.pdf.

MTN Ghana website, accessed November 2013, http://www.mtn.com.gh.

Osterwalder, A. and Y. Pigneur. 2010. Business Model Generation. New York: John Wiley & Sons.

Pelen, Margaux. 2013. “What a $70 Smartphone Means for Mobile in Africa.” https://medium.com/ what-i-learned-today/99674f8d4f6f.

Research ICT Africa. 2010. “Household and Small Business Survey.” Confidential Report (commissioned by Intelecon).

Research ICT Africa. 2012. “Household and Small Business Survey.” http://www.researchictafrica.net.

Stemle, Cary. 2013. “Direct Carrier Billing: The World’s Most Popular Mobile Payment.” Mobile Payments Today (blog), October 15. http://www.mobilepaymentstoday.com/blog/11377/Direct-Carrier-Billing-The-world-smost-popular-mobile-payment-Infographic?rb=false.

Tigo Ghana website, accessed November 2013, http://www.tigo.com.gh.

VisionMobile. 2013a. Developer Economics Q3 2013: State of the Developer Nation. http://www .developereconomics.com/reports/q3-2013/.

VisionMobile. 2013b. Developer Economics 2013. Developer Tools: The Foundations of the App Economy. http://www.visionmobile.com/product/developer-economics-2013-the-tools-report/.

World Bank. 2012. Global Financial Development Database (GFDD), World Bank, Washington, DC. http://econ .worldbank.org/WBSITE/EXTERNAL/EXTDEC/EXTGLOBALFINREPORT/0,,contentMDK:23492070~pagePK:64 168182~piPK:64168060~theSitePK:8816097,00.html.

32

9064_Ghana.indd 32

8/26/14 10:04 AM

9064_Ghana.indd 33

8/26/14 10:04 AM

Annex A. Mobile Applications

In this section, each of the mobile applications is listed along with their distribution channel, payment facility, platform, and revenue model.

While the various OS app stores constitute a major distribution platform, 7 of the 11 apps highlighted bypass formal app stores and are distributed independently. Eight of the 11 use formal banking facilities—that is, payment does not occur in-app but separately. The revenue model for 7 of the 11 is SMS-based content or services.

Farmerline Farmerline is a mobile-based service that provides agricultural information to farmers. After initially providing agricultural extension information by SMS, feedback from extension workers explained the low usage figures to be a result of illiteracy and the lack of detailed information in the SMS format.

While attending an ICT4D conference, Farmerline realized that there were lots of other countries that also needed a voice solution but only had SMS available to them. As a result, Farmerline developed four products: agricultural extension information to farmers via IVR, incoming voice calls from farmers to extension workers, mobile demand surveys, and SMS-based notifications by nongovernmental organizations/companies to farmers. The voice application includes market prices, weather forecasts, the ability to call in and get hold of an extension worker, and also a mobile survey tool where donors can measure the impact of their programs. Farmerline currently has around 100 farmers in Ghana and is facing difficulties in persuading farmers to pay for its products. As a result, Farmerline is adopting a new strategy: to focus on matching transactions between restaurants and suppliers. Its first target sector is the fish sector and they are using Google Trader as their platform.

In exchange for linking bulk buyers to fishermen, Farmerline takes a small monthly subscription fee of $2.5 from both the fishermen and the restaurants. Farmerline is primarily funded by two USAID projects, paying $2,000 per month for six months and $50,000 for three years.

Leti Arts Leti Arts is an African media company that builds mobile application games that feature African traditions and heroes. It develops games across all platforms, including Java, Android, and iOS. Its best selling in-house game is iWarrior, available at the iTunes store. Leti Arts sees games as a way to increase cross-cultural dialogue and increase interest in education. It developed Haki 2, a Kenyan game where players have to prevent the destabilization of the country by searching for clues in puzzle images and quizzes. Commissioned

by Afroes (www.afroes.com), it has over 300,000 downloads. Every game is backed by a digital comic, which is also for sale. The comics are available on the same platforms as the mobile games. Leti Arts charges 0.3 cedi per page and 4 cedis per comic and, depending on the platform, payment can be made by premium SMS, mobile money, or credit card. Premium SMS is the most popular form of payment because it reaches the largest group of people. In 2012, Leti Arts made revenues of $130,000.

In order to host the various games and comics that it develops both in-house and for other companies, Leti Arts launched Leti Centre in October 2013 (available for both feature and Android phones), which is essentially an app store. Leti Centre enables Leti Arts to provide a range of different games and comics within one application and to build its subscriber base outside the control of telecom operators.

In addition to developing games and comics, Leti Arts also does contract work, using African traditions and themes to build, for example, employee orientation games, as well as a game

34

9064_Ghana.indd 34

8/26/14 10:04 AM

TABLE A.1: App Summary Distribution Channel

Payment Facilities

Saya

OS app store

iWarrior (Leti Arts)

Platform

Revenue Model

Mobile money Formal banking facilities

iOS Android Feature phones

Advertising In-app advertisement

OS app store

Formal banking facilities

iOS Android Feature phones

App purchase Other

JumpFon

Independent Operator app store

Premium SMS

SMS/USSD

SMS-based content or services USSD/SMS-based content or services

BoxBuzz

Independent

Formal banking facilities

SMS/USSD

SMS-based content or services USSD/SMS-based content or services

Infoline (Nandimobile)

Google Play Mobile Operator VAS channel Mobile web / web OS app store

Premium SMS Formal banking facilities

SMS/USSD

SMS-based content or services Subscriptions services USSD/SMS-based content or services

mPedigree

Independent

Formal banking facilities

SMS/USSD

SMS-based content or services USSD/SMS-based content or services

Farmerline

Independent

Formal banking facilities Premium SMS IVR

SMS/USSD

Subscription SMS-based content or services USSD/SMS-based content or services

Esoko

Independent

Premium SMS

SMS/USSD

SMS-based content or services USSD/SMS-based content or services Freemium

mPawa

Independent Mobile web / web

Formal banking facilities

Feature phones SMS/USSD

Non-app related revenue Other

mNotify

Independent OS app store

Formal banking facilities

Android SMS/USSD

SMS-based content or services Subscription USSD/SMS-based content or services

Android

Advertising

Mobile money Trokxi

OS app store

None

In-app advertisement

35

9064_Ghana.indd 35

8/26/14 10:04 AM

FIGURE A.1: Business Model of Farmerline

Farmers

Operators

IVR, SMS, call to extension worker

Revenue share SMS and IVR

Monthly subscription

Monthly subscription

Fishermen

Google Trader

Restaurants

FIGURE A.2: Business Model of Leti Arts

Premium SMS Institutions Companies

Contract work Advertisement

Leti Arts & Leti Centre

Comics OS store purchase Mobile money

m-Apps

FIGURE A.3: Business Model of BoxBuzz

Subscription Post office

for the United Nations explaining developmental milestones for new mothers.

BoxBuzz BoxBuzz is an SMS application that provides a notification when mail has arrived in your postbox.

Consumers can avoid a trip to the post office to find out if mail or parcels or registered letters

SMS notification Postbox user

have arrived. The service is free to consumers; the post office or courier company pays a monthly subscription fee. BoxBuzz is not only targeted at post offices: it can also be used to manage pigeon holes in offices, schools, and apartments; for parcel deliveries at bus terminals; and for business parcel deliveries.

36

9064_Ghana.indd 36

8/26/14 10:04 AM

FIGURE A.4: Business Model of Trokxi

Advertisement

Free App

Businesses

Trokxi Trokxi is an Android application that provides estimates of fares based on the user’s location and destination. It provides a map route, a currency converter, some local taxi talk, and information on interesting local places to visit, based on the user’s stated destination. Users can upload new transport fares to the app. The fare estimates vary according to the time of day and the congestion rate. Trokxi is currently not earning any revenues,

Customers

but as it builds its user base it hopes to become attractive to advertisers.

JumpFon JumpFon is a premium SMS service offering a range of products such as SMS trivia games, mobile classifieds, health advice, and news updates. To subscribe, the user sends an SMS to a short code. The cost of receiving content is 0.20 cedi per SMS.

37

9064_Ghana.indd 37

8/26/14 10:04 AM

9064_Ghana.indd 38

8/26/14 10:04 AM

Mobile at the Base of the Pyramid: Ghana

Growing Innovation

Mobile

9064_Ghana_Cover.indd 1

Š2014 infoDev / The World Bank | 1818 H Street, NW | Washington DC, 20433 Email: info@infoDev.org | Tel + 1 202 458 8831 | Twitter: @infoDev www.infodev.org

8/26/14 10:08 AM