Introduction

This research-based dissertation sets out to establish the factors which affect the ability of broadband Internet businesses in the Developing World to be viable.  This attempts to answer the problem that is frequently observed:  Internet connectivity is established as part of a Development project, funded by grants from donors such as USAID or DfID.  When the project fails to generate sufficient revenues from users to fund ongoing costs, and donor funding ceases, the project also ceases.

A basic hypothesis behind the research is that if a profitable business model can be demonstrated, and it proves to be replicable, long term good quality connectivity can be provided without any donor funding.  If the business model is proven to be replicable, then such a model may aid the diffusion of Internet access and make a significant contribution to bridging the Digital Divide.

Why Broadband?

As the bandwidth commonly available to consumers in the Industrialized Nations increases, so published content and applications on the Internet are getting more and more bandwidth-hungry.  For example, the front page of www.apple.com decently featured the PC vs Apple Mac TV advert movies, which loaded by default.  On the slow Internet connections commonly found in developing countries, increasingly bandwidth hungry content will become increasingly unusable, and users may be restricted to whatever content and applications still work on slow connections.  The other side of this coin is that any business which makes significantly faster connectivity available will enjoy a substantial competitive advantage.

Defining viability

The dissertation discussed the definitions of sustainability and viability. “Sustainability” is a terms which is sometimes abused in Development circles.  In particular, repayment and Return on Investment of the initial capital is sometimes neglected, and ongoing maintenance / refurbishment / upgrading is not always adequately considered.

“Viability” is defined in a business as meaning an enterprise which can survive long-term without any advantageous funding or external support other than that which would be available to any business within the same economy.  This means that the cost of finance and expected repayment of loans / return on investment is at market rates, reflecting local economic conditions and risk factors.  Proper provision is made for ongoing investment to ensure that the products and services offered remain competitive.

The basis of the research

A particular location was chosen for the research: post-tsunami Northern Sumatra in Indonesia.  Initial research pointed to the selection of VSAT as the connectivity technology of choice, since the available dial-up and wireless services were considered to be overpriced and provided poor performance.  A business model was designed which would consider economies of scale and efficient use of available bandwidth to optimize revenues.  This was achieved by using half of the bandwidth purchased in an Internet café and half re-sold to local NGO’s and other businesses (delivered using point to multi-point WiFi.) The connection was designed to be the fastest in town.  This proved to be a significant competitive advantage, drawing plenty of customers even though access was priced 20% above the average.

Working with a long established NGO in the region, investment capital was found to implement the proposed business model.  After operating for six months, audited accounts were available.  The research was therefore designed to examine the business and see if the assumptions made during planning were validated by reality.  In addition, other factors where examined to see if there were additional lessons to learn.

The research comprised the following:

• A study of Internet café user demographics and their preferences.  This was achieved using a self administered multiple choice questionnaire, available in English and Bahasa Indonesian.
• A study of actual usage patterns, using monitoring software to record what applications users selected and which Internet resources they visited.
• Measurement of equipment utilization over time, also established using monitoring software.
• Measurement of bandwidth utilization over time, directly measured by the network router.
• A study of audited profit and loss statements for the business over the first six months.

Research findings

• In the sampled Internet café user population:
• 88% were educated to the equivalent of High School, and nearly 49% held at least an undergraduate degree.
• Only 4% of users think their family income is lower than average.
• 70% of users were male.
• 76% of users were in the 16-30 age range, with 59% between 16 and 25.
• 94% of users were Indonesians.
• Previous experience with PC’s and the internet were almost the same: Approx. 46% had more than 3 years experience.  About 91% had more than one year’s experience.
• 78% had access to PC’s at locations other than the Internet café.
• 44% had access to the internet elsewhere, but only 6% had other access to “fast broadband.”
• The traveling time to the Internet café averaged 15 minutes from home and 22 minutes from the place of work.
• 18% of users own their own business (about half of these employ others too.) 15% are unemployed. 25% are students.  The remainder work for NGO’s, local businesses or the government.
• 82% of users own a mobile phone.
• 74% of users visit the Internet café at least twice a week; 46% at least three times.
• 51% of users say they spend at least 4 hours a week at the Internet café. 28% spend more than 7 hours (7 hours a week costs the equivalent of around 1.5 times the average daily wage!)
• The factors which most influence the users’ choice of Internet café are:
• Connection speed (most important by a large margin.)
• Connection reliability
• Availability of air conditioning
• Good lighting.
• Connection cost was the eighth most important influence factor in the users’ choice of Internet café (out of fifteen.)
• Actual Internet usage (proportion of time spent) is:
• 62% browser
• 19% instant messaging client
• 7% media player
• Actual browser use (proportion of time spent) is:
• 21% web based email
• 25% general surfing and research
• 19% social networking sites
• 12% pornography
• About the Internet café business:
• Internet café utilization can easily reach 60% out of every 24 hours when everything is working reliably. This was seen in practice on a number of days. The business reported around 49% averages over the couple of months prior to this research.  During the period which was monitored, there was a higher incidence of technical problems, resulting in a 36% average utilization.  The difference between 36% and 60% utilization is $42,000 a year in revenue on the 30 computers in this business.
• The idea of deliberately splitting the utilization of bandwidth between mainly office hours users (NGO’s and other businesses) and Internet cafés (peak usage is late evening through to very early hours) works very well, and contributes to the optimization of revenue. This approach can make good quality bandwidth affordable for NGOs.
• Based on the assumptions adopted for cost of finance, pay-back and return on capital and service pricing, the following was calculated:
• Capital requirement for an 80 user Internet café is approximately 50% more than for 20 users.  The cost per seat is therefore much lower for a bigger facility.
• Scale and utilization levels are very significant in their affects on Profit and Loss.  It does not seem likely that a site of less than 40 users, with at least 50% utilization, can make a profit.  At 80 users and 60% utilization, the profit is a very healthy $180,000 per annum.

Implications

Demographics

The demographics of an area can significantly impact the potential success of  an Internet business.  Better educated people are probably more likely to be users, and since there is often a correlation between education and income, they may also be better able to afford it.  Conversely, illiterate people are unlikely to use the facilities.  Affluence also affects factors such as mobility.  People who own motorcycles will probably travel further to use an Internet café than those who must walk or cycle.  Thus, widely available transport creates a geographically larger potential market.

In the research location, there was clearly an opportunity to generate more business amongst women. In different cultures, there may be different factors affecting the gender balance of users, but in Sumatra, it was thought that maybe having a different room for women might have helped to attract more.

Optimization of costs

There is still some way to go in optimizing costs. In some locations, the management and costs of electrical power supply is very challenging.  It is thought that the use of thin client terminals and flat screen monitors can significantly reduce power requirements.

The manageability of the network is also a significant issue.  Drive-by installs often associated with pornographic and hacker/cracker websites can make computers unusable until they can have their hard drives re-formatted and re-loaded.  This creates work and reduces utilization levels.  Two obvious solutions are:

• Content filtering at source, making it hard for users to visit sites that often cause problems.
• Avoidance of susceptible operating systems. Possibly a change from Windows XP to a Linux distribution such as Ubuntu would be helpful.

Optimization of revenues

Further optimization is envisaged in a number of areas:

• Better utilization across the whole 24 hours of each day.  5am-11am is typically under-used.  Response times will be therefore be best during this period.  Utilization during this period should be encouraged, and possibly at a premium price so that it does not cannibalize income at other periods.
• Deliberate targeting of female users since their gender is under-represented.
• Differential pricing of standard and premium access speed – possibly including better specification computers for premium service too, along with different pricing at different periods of the day.
• Added value services such as training and “yellow pages” style online advertising.

Transferability

The research has allowed a reasonably good understanding to be gained of what works in a specific economy.  It would be worthwhile repeating the research in other locations, to determine if the conclusions drawn about the underlying business model studies apply generally. 

Post-disaster situations

This research would seem to suggest that a transition from donor funded Internet connectivity in post-disaster situations, to privately funded for-profit businesses, may be accomplished relatively quickly after the disaster event.  In the USA, this has implications for:

• Department of Defense, as it considers both Civilian / military liaison and so called “SSTR” (Security, Stability, Transformation and Reconstruction) operations mandated by DoD directive 3000.05.
• OFDA / State Department, as it considers the length of time that communications should be funded for organizations like UN OCHA HIC.
• USAID as it considers its part in accelerating recovery and development.  In addition to post-disaster scenarios, it could be helpful to be able to test the potential viability of enterprises that include an Internet connectivity component before committing funding. Re-engineering such business plans before projects start might significantly increase their impact and chances of success.

Viability prediction

The business model used in this study has been embodied in a spreadsheet.  With further validation, this could be a useful planning tool.

Some further work on the development of some demographic and economic indicators may also allow potential success and failure to be predicted.