Top Story

This week sees the publication of African Telecommunication Indicators 2004 by the ITU which is timed to coincide with Telecom Africa being held in Cairo next week. For a continent that all too often seems bereft of reliable ICT statistics its arrival is always heartening. The headline stories have been about the massive growth of mobile on the continent. But the report looks at more interesting questions including: future growth scenarios, actual levels of usage and return on investment. Russell Southwood digs below the surface to see what answers are being offered to these questions.

At the end of 2003, six Africans out of a hundred had a mobile phone, twice the number that had access to a fixed line: six million subscribers in all. Nevertheless this remains the lowest mobile penetration of any region in the world. The report puts forward three scenarios for growth that will be achieved by 2010: low (10.2 million, a 70% increase); medium (15.3 million, an increase of 255%) and high (20.4 million, an increase of 340%). These figures might seem fanciful but as the report shows most of those making predictions in the past have under- rather than over-estimated the likely growth.

So why have those making predictions go it so wrong in the past? Well, the answer is that the income statistics for Africa - especially at the lower end of the scale - have proved to be a very unreliable predictor of mobile usage because of "informal markets and the unreliability of existing income data." In other words, there’s more money out there than the World Bank figures are showing:"The average African expenditure of USD27 per month for mobile service is more than half the average African per capita is difficult to imagine a subscriber (actually) spending such a large portion on their income on mobile services."

The lower the reported income, the less reliable the statistics become:"The link between income and mobile density for incomes less than USD500 a year is much weaker than for incomes greater than USD500 per year... National analyses of mobile demand reveal much higher levels of penetration than would be expected by income". What this is actually saying is that in countries as diverse as Nigeria and Kenya there is far greater mobile potential because there is more money in the economy than the available figures have identified. Nigeria is now estimated to have 15-20 million potential subcribers, double what was predicted two years ago.

If the average penetration is 6% overall then for all the staggering, vertiginous growth there are not many people who will be affected by it. Not quite true. At present 60% of the population on average is in area that offers mobile coverage and this percentage will (on the basis of our discussions with operators in key markets) go up to 65-70%. Also counting the number of mobile phones does not give an accurate picture of actual usage. The report cites two other reports looking at access to mobile telephony by household rather than by individual. On this basis 32% of households had access to mobiles in South Africa and 31% in Morocco. And it remains important for rural areas as 13.8% of households had access in Morocco as against 2.3% with fixed lines.

Admittedly these are two of Africa’s wealthier markets but getting to the bottom of how many people actually have access to using a device is key to making sense of what is happening to the digital divide. To address this issue the ITU is encourage the statistical services of African countries to include questions on this subject when they carry out census work.

So who are Africa’s mobile investors and what kind of return are they making on the money they have invested? The top six operators across the continent in order of scale of revenues are: Vodacom, MTN, Orange, Orascom, Celtel (formerly MSI) and Millicom. Interestingly those that are doing best tend to be focused in a relatively small number of countries, usually between 6-8. Will they do so well as they expand their field of operations? Strategically, almost all have chosen to avoid going head-to-head with competitors in all their markets.

On the basis of looking at returns in company accounts, profit as a percentage of revenue varies between 10.6% (MTN) and 16.6% (Celtel). The exception is Millicom which is showing a return of 42%. A word of caution needs to be entered. Many of these operators are still investing in their networks and these have not yet become mature markets. Investment yields will go up as markets mature which they must surely do by 2010 if one of the growth scenarios cited above turns out to be credible as well it might.

Africa’s mobile strategic investors
Top mobile groups in Africa by number of proportionate subscribers, December 2003

Note: * Subscriber data refer to year end; financial data to year ending 31 March. ** EBITDA instead of profit. *** MISSING

****While the total number of subscribers refers to all subscribers within the countries where the strategic investor has operations, the proportionate number of subscribers are those of the strategic investor. Example: The total number of subscribers in the 5 countries where Vodacom has operations amounts to 10’184. Out of these, 9’666 are Vodacom’s subscribers.

Source: ITU adapted from company reports.

Top ten mobile operators in Africa
Ranked by number of subscribers, 2003

Source: ITU adapted from company reports.

On slightly older data (2002), the top ten mobile operators operators are showing Average Revenues Per User of between USD11 (Maroc Telecom) and 30% (Cell C - South Africa). Exceptions are USD69 (MTN Nigeria) and USD65 (Djezzy - Algeria). The Nigerian figure only goes to reinforce the point made earlier that Nigeria is a richer country than the official statistics show.

Where are the areas for future revenue growth? The world average for SMS text messages is 4 per subscriber month. But in many countries, especially among the young, this cheaper, thumb-operated method of communication has almost taken on the status of a major virus. Relatively speaking, Africa is thus far behind the wave but the signs of "catch-up" are there. Again the data is one year behind but the report shows 17 SMS per month in South Africa and 11 in Mauritius. Ah, but I hear the sceptics carping, these are again the wealthy countries of the continent. Yes but...A more recent breakdown of SMS use by subscriber type shows prepaid subscribers averaging 9 per month in South Africa. Furthermore Togo (on the basis of those 2002 figures) was averaging 6 per month and Cote D’Ivoire 4 and Cape Verde and Burkina Faso 3.

The report also makes the case for mobile internet on the basis that people will use mobiles to connect their laptops. This will all rely on operators being able to offer something above the GSM standard of 9600 kbps. Operators in wealthier markets (Egypt, Morocco and South Africa) have already launched GPRS networks but these are unlikely to be rolled out in less wealthy markets. For example, operators in Nigeria have spent (or will spend) USD200 million installing 3G standard CDMA 2000 1x networks for fixed wireless access. One of the Nigerian operators is offering internet access speeds at what it claims will be up to 144 kbps. However this type of network offers limited mobility which may suit certain types of local users but not others.

There are also a number of other wireless technologies that might impact on the problem. However as they will tend to cannibalise the revenues of the most probable investor - the incumbent telco - it may well be some time before they are implemented.

In our view, e-mail to mobile for the receipt of messages is a service waiting to take off. The mobile is Africa’s device of choice and both texting and e-mails are used partly for that most compelling of reasons: cost. One operator recently reported 50,000 people trying a service of this kind in its first week. The main impediment is whether the mobile user’s phone has the capacity to handle the traffic.