January 2008
The availability of low-cost mobile phones and the already broad coverage of GSM networks in Developing Countries are a huge opportunity to provide services, based on Information and Communication Technologies (ICT), that would trigger development and improve people’s lives. A first step in that direction has been the apparition, last two years, of numerous success stories using mobile phones in development. However, there is still a gap between the development of few services that demonstrate the proof of concept, and the availability of thousands of services in all Developing Countries of the World. This paper explores this gap, and the potential solution to fill it with the enabling of Mobile Web technologies.
In a first part, this paper revisits the current technologies, mostly SMS, used in the mentioned success stories, and their abilities to scale well. In a second part, we present the strength of Web technologies, the major challenges specific to this context, and some potential directions to address them. In the last part, we introduce a potential program W3C may start to help enabling the next generation of Mobile Web applications for rural communities of the Developing World.
The emergence of new Information and communication technologies (ICT), the Web and Internet in particular, in late 80s, has changed the World, offering a new paradigm in communication, exchange and commerce. However, while the new Information Society is still developing today, a new gap has also appeared with those without regular, effective access and ability to use these digital technologies. This is known as the Digital Divide, which is particularly affecting Developing Countries.
On another hand, ICTs are also a great opportunity for the Developing World. Providing minimal services (Health, Education, Business, Government...) to rural communities and under-privileged populations is of major importance to improve people lives, and to sustain development. Using ICTs would be the easiest and possibly only way to develop and deploy those services. It is therefore critical to work towards bridging this Digital Divide.
In this context, the recent explosion of mobile telephony in the Developing World is a great opportunity. At the end of last year, according to the GSMA and ITU, the total number of people having accessing to a mobile phone was around 2.7 billions, and 80% of the World population was currently covered by a GSM network (source [1], [2]). These numbers illustrate the potential of the mobile platform to be the right solution to deploy services now, compared to other options, which are still in development phase (e.g. low-cost laptops). The first measurable results are even already available. A recent study of the London Business School demonstrated the impact of mobile phones and associated services on productivity and social development, showing that 10 more mobile phones per 100 people increase the GDP of a Developing Country by 0.6 percent (source [34]).
However, the potential is far greater than this number. Indeed, it is still quite hard to develop, and widely deploy reliable eServices on mobile phones, targeted at specific communities’ needs.
This paper will review, in a first part, the strengths and the weaknesses of existing approaches used to deploy eServices on mobile phones in Developing Countries, and will emphasize, in the second part, on the Mobile Web option, its potential, and the associated challenges. The last part will be dedicated to present the next steps W3C ([3]) is planning to implement to address these challenges.
For about two years, the use of mobile phones as a platform to deliver services to underprivileged populations and rural communities has been experienced all over the Developing World, particularly in India and some countries of Africa. Many stories have demonstrated the ability for ICT-based applications to improve people lives and increase their incomes with simple services (read e.g. [4]).
The type and domain of deployed services are very broad, and spanning the whole society. There are lots of examples in the banking area (e.g. [5]), business (e.g. [6]), health and government / public services. Usually, each service has been the result of a specific project directed by a local or an international NGO, or through a governmental experiment .
From a technological point of view, most of the projects are using SMS technology, and some, very few in fact, are voice-based (see e.g. [7]). This section will present in details how each type of applications works, and what the strengths and the weaknesses of each technology are to cope with the specificities of the targeted user community.
Today, SMS applications are the most common way used in the Developing World to provide eServices. People, knowing the phone number associated with the service, send an SMS to this number with appropriate keywords, and get back the answer by SMS. Sometimes, when there is no interaction needed (e.g. sending weather forecast), it is just a broadcast of SMS messages to people subscribed to the service.
The reasons of the success of this technology are numerous:
However, there are also some weaknesses associated with this way of providing eServices. These limitations can be classified in two sets, those inherent with the use of textual information for both the input and the rendering, and those specific to the SMS technology.
Concerning the use of textual information, there are 3 major problems:
Out of these limitations which are common to all textual approaches, there are also specific weaknesses inherent to the SMS technology:
While SMS-applications are clearly the most used techniques to offer eServices today, we think that this is due to the lack of better solutions. Given the mentioned intrinsic weaknesses, we can’t see how it could be possible to deploy a big numbers of services at a large scale, targeting a population of millions of people.
Unlike other types of applications, Voice applications don’t have any specific module on the mobile phone. People are just placing a traditional phone call at a specific phone number to reach the voice platform from which the service is accessible. From there, navigation through the application is done either by voice input (the user speaks to the application) or by pressing the phone keypads. This type of applications has some specific strengths:
Despite these specific features, there are also issues that may delay the adoption of voice technologies in a near future:
Eventually, Voice applications clearly have the right potential to help in some specific areas, particularly to tackle the literacy problem. However, particularly in the speech synthesis domain, further research is necessary before considering voice technologies as the recommended way to provide eServices. Even if the technology was appropriate, such applications would require investment at a country level, and the complexity of application development will still be a barrier to enable local people. Like for SMS, Voice applications are a potential solution in some specific projects, but we don’t believe that it has the potential to be the right solution to provide a huge number of services all around the World.
While all the existing success stories demonstrate the potential of mobile phones to be the most promising platform to deliver services to rural and under-privileged populations of Developing Countries, we doubt that the current technologies would allow a large scale development and deployment of applications. Moving from “islands” of successes as it is today, to a World where thousands of services are available, usable and useful for under-privileged populations is a huge challenge which could hardly be tackled by SMS or Voice applications. We believe that only the switch to Web technologies will allow this transition, and we will present, in the next section, the roots and rationale of this view.
The Web is clearly an incredible space of communications and exchange as well as an endless source of information. For that reason, W3C, the standardization organisation for all Web core technologies ([3]), has a mission to work toward Universal Web Access (the Web for Everyone, at Anytime, from Everything). Enabling Web access from mobile phones is part of this mission, and is the specific work item of the W3C Mobile Web Initiative ([19]).
However, providing access to the Web may not be a goal by itself for rural communities. The aim is to provide services to populations using the power of the Web as the support for facilitating the development and the deployment. Among its numerous features, the Web has indeed some specific strengths in this context:
These specific strengths let us think that Web Technologies are the most appropriate way for future large scale, low-cost development, deployment and availabilities of ICT-based services on mobile phones.
However, there are very few success stories today about Web usage on mobile phones in Developing Countries. This is due to the existing challenges requiring appropriate adaptation of existing standards, and new ways of developing content and applications. We will explore in the next section of the paper what are these challenges, and which actions could be engaged to tackle them.
The strengths of the Web are obvious, and its incredible success in the last ten or twelve years illustrate them. The recent take-off of Mobile Web access in the Developed World is also an evidence of the importance of Web access on mobile phones. As underlined in the section 2 of this paper, the potential of the mobile platform in development is also clear. The question is to know if these two aspects fit together, and if the Mobile Web is a promising opportunity for Developing Countries.
Lots of people are indeed doubtful on the potential of the Mobile Web to be a solution to improve people lives, and leverage development of rural communities (read [20], [21]).
We will review in the following the usual problems people are considering.
Lots of people think that data service and web-capable handsets are very expensive, and rural communities would never be able to afford them for Web browsing. This is true, if we consider only the cost of the access. The critical aspect is the return on investment. It is very unlikely that a crop producer in rural India would spend the required money to surf the Web for entertainment, because he would consider the money wasted, and not invested in his basic needs. At the opposite, if accessing a service to declare his new children costs him a day of salary, but the travel to the nearest office would require more than a day, then he would surely go for the online version. This specific aspect has been very well studied in the literature (see [4] and [22]). Moreover, the overall model of the mobile telephony, concerning the voice service, is here to illustrate these theories. 4 or 5 years ago nobody would have bet that Developing Markets could be a business opportunities for Mobile players, and now they are considered as the most beneficial ones. All the models which worked for the voice service (community shared access, phone ladies model [23]…) are able to work for data access too. So the cost of data access should not be a problem only if valuable content and services are available to potential users.
If we look at how the Mobile Web has been evolving in Europe, US or Japan last 2 or 3 years, we can see that the trend is to aim at providing the same user experience as on a classical desktop machine. Device manufacturers, operators, browser makers… all mobile players are working in that direction. Given that Web content in 2007 are highly multimedia, with video, sound, picture and so on, high capacity networks and powerful devices are indeed required to access those. Are these requirements the same in our context? Probably not. One would not need sound and video, or color screen to provide useful services to community. And therefore, GSM should be perfectly ok to deliver quickly that kind of services. Again, the success of the SMS applications illustrates this aspect. Unfortunately, as of today, all the low-end devices, targeted at Emerging markets don’t have a web browser. Such capability appears usually in the expensive devices. This is clearly a problem, and a potential limiting factor. But is it technically challenging or costly to provide basic browsing capabilities on low-end phones? We tend to think that it is not. Indeed, what is really required is at least a minimal browser being able to render xhtml basic, css mobile, and probably also a simple image format. Late 90’s, all mobile phones had such minimal capabilities with the availability of WAP browser. It was not Web standards, but a specific markup (WML), but it was very close to xhtml-basic. So having a similar capacity today on all phones is not a hard thing to do, is not technically challenging, and does not require a very powerful handset. The current lack of such a minimal browser is most probably due again to the lack of demand from users, because of the lack of useful content.
As we can see, the critical problem is to make useful content for rural communities, content that would help people in their daily lives, and allow them to get more income and afford the service. In the last part of this section, we will focus on defining what is a useful service or useful content.
Before being useful, a service or content has to be usable. The usability aspect, heavily explored in the application design literature (see e.g. [24]), is particularly critical when developing services towards users without technological background. Indeed, the users targeted by our work are clearly people of rural communities in Developing Countries. Those people usually present two specificities: no previous computer experience, and no technical background (in computer science, and computer-based application in general).
In the specific context of Web applications, the user interacts with both the browser, and the Web content loaded. Therefore mobile browsers and Web applications have to be accessible and natural for those without previous desktop experience, and without technological background.
Concerning browsers, there are different issues to consider. Usually, the configuration is the first problem. While placing a phone call or sending/reading an SMS is almost obvious on all handsets for almost anybody in the World, configuring a terminal to access the Web is very far from being natural and easy. Only expert users can do that configuration themselves directly. So given that this is already a problem for most of users in Europe or US, it is hard to believe that this would not be also a problem of a higher degree of magnitude in Developing Countries.
Then, the interaction with the browser is also a problem. All the usual menus and interface have been defined by reproducing what exists on usual desktop applications. The notion of URI is also less than natural. It is not possible to believe that people would enter a meaningless string of characters by hand, through the keypad.
To end, the way of rendering Web content has also to be redefined. As said earlier, the current trend in the Developed World is to focus on providing the same experience as on desktop. For instance, towards reaching this goal, browser makers are developing zooming interface ([25]). These kinds of interfaces are very successful to allow anyone to access on a small screen content which is defined for big desktop screen. But are such zooming interfaces natural for those who never been on the Web before? This is questionable. A new way of interactions with the browser should probably be defined, and there are ongoing researches in that direction (read [26]).
Concerning Web content itself, the problem is identical. There are some challenges that are related to the specificities of the mobile platform. These challenges are currently addressed by the W3C Mobile Web Initiative ([19]) and are not specific to context of this paper.
There are also specific challenges inherent to the targeted users. It is essential to understand what is natural, in terms of interaction between the user and the content for those who never used ICT-based applications (read [27]). Here again there are few ongoing researches in that area, but not related to the specificity of Web content. It is therefore critical to revisit these researches and their applicability.
To end, there are also specific challenges on content accessibility which are currently addressed by the W3C Web Accessibility Initiative ([28]). However, further work is needed concerning the problem of illiteracy which is particularly affecting Developing Countries.
While offering usable content and browsers is a mandatory step to make the Mobile Web relevant, the major goal is still to provide useful applications which would really improve people lives. The question is to understand which (type of) services or content would be considered as useful by targeted populations.
It is very unlikely that we could define a list of applications that could be considered useful by all, all over the World, from Kenya to India, or from Bolivia to Sudan. Of course, there might be some particular examples. M-banking is surely one, because it is a common process worldwide (exchanging money) and therefore making such a service available to all is very important. But for other services, there are probably big domains to tackle (education, health, public/government, business…), but trying to define a list of services to provide to communities would surely not work (see e.g. [29]).
At the opposite, if we analyze the success stories we mentioned before, we can see that they are sharing the characteristic of enhancing existing behaviors or way of working or living. Attempts to radically change the habits are usually not working. All the successful projects are usually the result of in-depth analysis of a community, with the involvement of local NGOs working with the community for a long time.
That said, there is probably a place for general guidance. Each project has developed its own expertise, but there are surely commonalities, and a way to define some guides that would help the identification of the areas where a service will be useful. It should be possible to gather expertise which has been developed through the realization of existing projects, plus theoretical knowledge from e.g. the ethnology field, to develop guidelines on identifying areas where ICT-based services could help and leverage development of specific communities of the Developing World.
The case of public/government services is notably different. It is not up to a specific community to define the government services. This is up to the government bodies to analyze the mandatory interactions between the citizen and the state and then provide those services to the communities to ease their communications with the state. If Government working on eGovernment framework wants to take advantage of the eServices to leverage and facilitate the interactions and relationships with rural communities, and if they want that to make their program successful, they have to take into account the specificities of the mobile channel, and the way to make usable services as described in the previous sections, and that at the early stage of requirements, and not as an add-on at the end of the program. W3C is currently starting to explore the area of eGovernment framework (see [30]) and will take into accounts the specificities on the Developing World, and the case of M-government as described here.
To summarize this section, we don’t believe that rural communities of the Developing World would find today a real value to access existing Web resources from their mobile phone. However, Web technologies on mobile phones are a great opportunity to deploy a big numbers of application at large scale and low-cost, enabling anybody to start and run new services. In order to make this potential a reality in a near future, specific work has to be engaged, particularly in the area of Human-Computer Interaction (HCI) applied to the targeted users, and in the area ICT for development (where and how ICT can leverage development).
As underlined in the previous section, to make the Mobile Web relevant, usable and useful for rural communities of the Developing World, it is essential to engage work in different directions:
Obviously, those broad directions have to be refined to establish detailed and achievable goals and deliverables in a short and realistic timeframe. W3C will propose the creation of a specific Interest Group ([31]) to address the definition of this roadmap.
In order to be successful, such initiative should also integrate some other critical steps we will explore in this section. Indeed, the success of the approach, like for any work at W3C, will rely on the participation of key players, and essential stakeholders. Here, we need to engage different communities which are not yet working together:
The involvement of representatives from all these communities is essential to ensure that all aspects of the problem will be tackled.
Another critical step to take into account is the dissemination. As mentioned in a previous section, a huge number of useful services will effectively appear only if lots of people know how to develop them. It is important to work on the appropriate technologies, but without empowering people, it would have no effect. In our context, the targeted audience is clearly IT sectors of Developing Countries, students and future actors of these countries too, and people from all over the World who are interested to work in the development sector. It is therefore critical today to work on setting up new curriculum at Universities of the Developing World that would teach students how to develop ICT/Web-based applications on mobile phones.
Concerning the problem of illiteracy, as mentioned before, it is obviously another key point to take into account in order to provide services to billions of people.
However, this is by itself a huge domain not specific to the Developing World, even if they are more affected. Clearly, here again, Web technologies are far more able to cope with this issue than e.g. SMS applications. There are ongoing researches exploring for instance the use of meaningful icons (see [32]). W3C, in the Web Accessibility Initiative, is aware on the importance of this issue (see [33]), and this is a subject we will focus on in the future.
Clearly the mobile phone and the availability of GSM networks in Developing Countries are a great opportunity to bridge the Digital Divide, and to provide minimal services to rural communities. While the proof of concept has already been made through numerous success stories, the technology used so far, SMS, will not allow a large-scale, low –cost development and deployment of services. Today, only the use of Web technologies on mobile phones would help reaching this goal. However, some adaptations are required to make the Mobile Web relevant and useful for the targeted communities. This concerns mostly the interfaces between people without prior experience with ICT and without technological background, and the applications (mobile browser, and Web content and applications). It is also critical to develop expertise and guidelines on how to identify the exact needs and requirements of specific communities in terms of ICT-based services. Eventually, a particular attention should be set on empowering people and enabling them to use the materials we will develop.
W3C is planning now to launch a specific Interest Group to focus on building the appropriate community with representatives from all parties, and define in details a roadmap that may be tackled by a new activity or initiative in a near future.
[1] Universal Access: How Mobile can Bring Communications to All, GSMA Report, October 2006
[2] Telecommunication/ICT Markets and Trends in Africa, ITU, 2007
[3] World Wide Web Consortium (W3C) http://www.w3.org
[4] R. Jensen, The Digital Provide: Information (technology), Market, Performance, and Welfare in the south Indian Fisheries Sector, The Quarterly Journal of Economics, Vol. CXXII August 2007 Issue 3
[5] M-Transaction, Economic Empowerment, Vodafone
[6] Safaricom, Getit411-Sokoni service
[7] MoSoko Project http://wiki.nrcc.noklab.com/Mosoko
[8] Grameenphone foundation, Village-Phone program, http://www.grameenfoundation.org/what_we_do/technology_programs/village_phone/
[9] SafariCom M-Pesa system http://www.safaricom.co.ke/m-pesa/
[10] J. Donner New collection of m-banking papers, June 2007 http://jonathandonner.com/archives/11
[11] W3C Internationalization activity, http://www.w3.org/International/
[12] Ethiopians get texting in Amharic BBC News December 2007
[13] Information Services and Access Mechanism of Mobile Web for the Under-privileged, Asoke K Talukder, December 2006, http://www.w3.org/2006/07/MWI-EC/PC/MobileWeb_daimlerchrysler.pdf
[14] Mobiles Phone banking takes off, Private
Sector Development Blog, June 2006
http://psdblog.worldbank.org/psdblog/2006/06/mobile_banking_.html
[15] busineSMS messaging tools http://www.businesms.com/2006/index.html
[16] Frontline SMS, kiwanja.net, http://frontlinesms.kiwanja.net/
[17] Voice Extensible Markup Language
(VoiceXML) Version 2.0, W3C Recommendation,
http://www.w3.org/TR/2004/REC-voicexml20-20040316/
[18] W3C Workshop on Internationalizing the Speech Synthesis Markup Language (SSML), November 2005, http://www.w3.org/2005/08/SSML/ssml-workshop-cfp.html
[19] W3C Mobile Web Initiative http://www.w3.org/Mobile
[20] K. Banks, the Digital Divider, Sept. 2007
[21] N. Eagle, the Mobile Web is NOT helping the Developing World... and what we can do about it, Nov. 2007
[22] I. Quadir: The power of the mobile phone to end poverty, Jul. 2005
[23] R.Buerk, BBC New, Nov. 2005: 'Telephone ladies' connect Bangladesh
[24] J.D.Gould, S.J.Boies, C.Lewis: 1991, Making usable, useful, productivity-enhancing computer applications, Communications of the ACM.
[25] Wikipedia, Definition of Zooming User Interface
[26] D.Joshi, V.Avasthi, Motorola: Mobile Internet UX for Developing Countries, Sept. 2007
[27] A.Maunder, G.Marsden, W.D.Tucker: Evaluating the relevance of the ‘Real Access criteria’ as a framework for rural HCI research, to be published as part of the CHI-SA 2006 conference proceedings, Cape Town, 2006. pp 75-79
[28] W3C Web Accessibility Initiative http://www.w3.org/WAI
[29] G. Marsden: Doing HCI Differently in the Developing World, May 2007
[30] eGovernment at W3C http://www.w3.org/2007/eGov/
[31] W3C Interest Group, World Wide Web Consortium Process Document, http://www.w3.org/2005/10/Process-20051014/groups.html#GroupsWG
[32] I.Medhi, A.Prasad, K.Toyama, Microsoft Research Labs India: Optimal Audio-Visual Representations for Illiterate Users of Computers, WWW2007 http://www2007.org/papers/paper764.pdf
[33] Social Factors in Developing a Web Accessibility Business Case for Your Organization http://www.w3.org/WAI/bcase/soc.html
[34] Leonard Waverman, Meloria Meschi and Melvyn Fuss: The Impact of Telecoms on Economic Growth in Developing Countries