Arthur Goldstuck, South Africa’s best-known ICT strategist and web researcher, has a difficult task. Every year, his World Wide Worx research organization presents its thoughts on where South Africa’s Internet user base is headed in the next few years. Of special interest are the Internet user statistics (and the factors that influence these numbers).

From 2003 until 2007 W.W.W.’s predictions were spot-on. After a period of rapid growth in the 1990’s, single-digit Internet user growth rates were the norm for much of the next decade. After all, the South African telecom industry was operated under a monopoly and this constraint gave researchers an easier time in forecasting trends. Telkom essentially controlled all factors that influenced South Africa’s Internet adoption. However, the market opened around 2008. Around the same time, broadband began to arrive in earnest as costs were lowered. Mobile Internet was soon to follow. Currently, undersea cables, the licensing structure for ISPs, the embracing of broadband by SMEs, and cellphone connectivity all contribute heavily to Internet usage trends in South Africa. Arthur Goldstuck’s forecast model must be as complex as ever.

Case in point: Two-year predictions made in 2008 low-ball actual Internet user numbers seen in 2010 by approximately 800,000 people, a 12% deviation. Again, in 2009, Mr. Goldstuck foresaw impressive growth of South Africa’s Internet user base, but not to the extent that South Africa actully is witnessing. That year, Mr. Goldstuck went as far to anticipate 9 million Internet users by 2014 – enormous growth given the low rates just years prior. A year later, however, Mr. Goldstuck dramatically altered his forecast by predicting 10 million Internet users by 2013 (an extra million users in one less year!). The reason: extreme mobile Internet adoption that Mr. Goldstuck now expects to exceed supply within a matter of years.

Now, the above example is not to fault anyone for under-estimating the power of the undersea cable, telecoms competition, the popularity of the mobile device, or any other economic factor in driving Internet growth. Instead, the take-away is that predicting Internet trends in developing nations is near-impossible.

Fortunately, World Wide Worx aptly addresses the larger picture in addition, even if exact numerical trends are tough to nail down. In 2008, the firm accurately judged that Internet growth would be amongst the upper and middle class segment of the population until the market reaches saturation after 2012 – unless the country finds a way to drive down costs. Three years later, those costs have dropped and the prediction remains true: massive “re-invention” of the industry will lead to supply outpacing demand.

Excerpts from the annual World Wide Worx reports:

2003 (exact):

• 3.1 million South Africans have access to the Internet at the end of 2002
• Growth in 2003 was set to be only 6%, with 3.28 million South Africans expected to have access to the Internet by the end of 2003 (Actual = 3.283 million)
• http://www.engineeringnews.co.za/article/sa-internet-access-expected-to-surge-2003-12-01

2005 (exact):

• 3.6 million South Africans were to have access to the Internet at the end of 2005 (Actual = 3.6 million)
• This means growth in 2005 ticked up slightly from 4% in 2004 to 5% in 2005
• http://www.worldwideworx.com/2005/12/01/great-expectations-stall-on-the-internet/

2007:

• The number of South Africans with Internet access will rise by just 3 percent in 2007 due to high tariffs, but stiffer competition in the broadband market will boost connectivity by 2010
• 3.85 million people in South Africa will have access to the Internet by the end of 2007
• http://www.worldwideworx.com/2007/07/05/sa-internet-access-growsbut-only-for-the-haves/

2008 (conservative):

• 4.5 million South Africans were now using the Internet
• Expected South Africa’s registered Internet users to increase by 13 percent in 2009 and as much as 17.5 percent in 2010 (Actual = 15% from ’08-’09 and an astonishing 28% from ’09-’10)
• http://uk.reuters.com/article/2008/12/04/safrica-internet-idUKL413781020081204

2009 (conservative):

• 6.1 million Internet users by 2010 (Actual = 6.8 million)
• The country’s number of Internet users could grow to nine-million by 2014 (Already at 6.8 million in 2010)
• http://www.engineeringnews.co.za/article/sa-internet-users-seen-growing-to-9m-by-2014-2009-03-26

2010 (conservative):

• Internet user base grew by 15% last year, from 4.6 million to 5.3million and a similar trend is expected for 2010
• By 2013, Goldstuck estimates a reach of 10 million internet users in South Africa. (In 2009 this prediction had been 9 million users by 2014 – one million fewer in one year longer)
• http://www.itnewsafrica.com/2010/02/bandwidth-capacity-in-sa-to-double-by-2011/

2011 (ambitious):

• The group noted a total internet user base at 6.8 million in 2010, from 5.3 million in 2009 and 2.4 million in 2000 (2008 prediction would have 2009 at 5.1 million and 2010 at 6.0 million)
• SA had six million mobile internet users, of which five million already had access to PCs
• The demand for mobile services will outstrip supply by 2013
• http://www.iol.co.za/business/business-news/tornado-effect-to-hit-sa-s-mobile-1.1048395