Four thousand people are slated to attend the industry’s biggest continent-wide talk-fest AfricaCom next week, which has been given the upbeat title of Driving the Next Stage of Telecoms Growth in African Telecoms. However, the mobile masters of the universe face tough challenges ahead with growing competition and the levelling off of market growth. Operators face a slalom of obstacles that will begin to separate those who have the skills from those who have simply been lucky to have been in the right place at the right time.
In the words of the apocryphal Irishman giving directions:”If you want to go there, I wouldn’t start from here.” African Mobile operators started out as narrow-pipe, voice operators at a point at which GSM technology costs had fallen to more modest levels. The pent-up demand for voice meant that this technology was well-used to meet that demand and make a fabulously good living.
Consumers in African countries were much more forgiving of the kinds of congestion and poor coverage than their counterparts in more developed countries. On his arrival at Safaricom, the CEO Bob Collymore said that he hoped to solve the network congestion levels in Nairobi, blaming it on the high density of phone users and the practice of “beeping”. In my memory, these congestion levels have been there for at least eight years and Safaricom is still the number one player so you get some idea of people’s tolerance levels.
As he himself said:” “I have been using the network and it has issues. We have a fantastic customer care service, but volumes have grown and strained capacity”. Getting through to the customer care help line can be about as hard as navigating the very considerable Nairobi rush-hour jams. I have sat and lunched with senior customer care executives from several companies in another country who told me a very similar story. You might tout customer service as your defining feature but getting it delivered is another matter.
But the company has to face these challenges in a the middle of a price war initiated by Zain’s new owner Bharti Airtel. It will undoubtedly initiate a similar price attack across all of it territories and has announced a succession of outsourcing contracts that will undoubtedly address the expenditure side of the balance sheet with some vigour, something that not all operators have yet got their heads around.
But with falling voice ARPUs and greater price competition, the challenge has been to drive up data revenues. This has meant deploying networking improvements (the alphabet soup of GPRS, EDGE, 3G and onwards and upwards) in order to capture the new data opportunities. And in this area, a number of mobile operators have been successful in driving data revenues up to 10-15% of overall revenues and still maintaining a good margin on the service.
In the best countries, a combination of the business power user with a laptop and 3G modem and the enthusiastic, young Facebook user has allowed operators to add revenues to replace some lost voice income. However, incomes will be much less rosy in those countries where literacy is low. A recent audience survey showed that 44% of mobile users in Ghana did not use SMS, the most basis of non-voice tools, and Ghana is an African country with relatively high levels of literacy.
Operators rarely break out voice and data in describing their ARPUs but this will become increasingly important information for investors in the years to come. For this transition from voice to voice and data goes through a cycle and initially prices are high and then cascade down towards the more reasonable as more competitors enter the market.
Lower prices bring in more users but the bottom of the price curve simply returns you back to low margins again. And all the time, growing data demand means continuous investment in networks that were not initially provisioned for data and the investment of considerable sums in national fibre backbones to accommodate this demand. So if the operator has to do this to stay in the game, why not also target the corporate customers, particularly the multinationals operating across the continent?
In a hop, skip and a jump, you have become a vertically integrated telco offering everything to everyone and not quite offering something to someone. At a conference in July in Nairobi, there was a revealing set of exchanges at a breakfast sponsored by Safaricom to pitch their services to a particular sector of the business community. After an articulate presentation of how the company could help business users, a number of people began to rip into them about customer care and service levels.
The responses from Safaricom staff were sufficiently honest and straightforward to make you realise that it is one thing for a vertically integrated telco to take the ground but quite another to actually then run all the different services effectively. Time may improve matters but the fundamental question remains: how do you concentrate on so many different things at once and do them well?
Furthermore, although the shift to data puts a spring in the step of most mobile executives, the shift to an interest in services and apps has the potential to marginalise them as “dumb pipe” operators. The new generation of OS operators (Blackberry, iPhone, Android and others) are offering services and apps in a way that the mobile operators failed to do.
Some of them may now be meeting behind closed doors in Europe to plot a fight-back (and Vodafone and Nokia may believe this is possible) but Africa’s mobile operators are fundamentally not “content guys”. Beyond M-Pesa and its copies, name a highly successful mobile app or service outside South Africa. I can already picture you not needing the fingers of more than one hand.
The truth is that the income splits in Africa are too disadvantageous to allow the mobile operators to do anything more than collect tolls for allowing people across the bridge. This is no way to develop the service and app business and unless something changes, those content operators will be off where someone at least offers them a modest margin for manoeuvre in terms of what they can do and how they are rewarded. There may only be 300,000 operating iPhone apps in South Africa at the moment but this is the start line and the race has just begun. SMS services traffics indicate far higher levels of interest.
Lurking at the edge of this picture is the possibility of a disruptive entrant in one of Africa’s more competitive environments; an insurgent challenger that changes the basis of the CAPEX and OPEX question by operating fundamentally cheaper technology. I have been suggesting this possibility for over five years (talking occasionally of “desktop GSM”) and its failure to arrive may be comforting to those who it would most affect.
Nevertheless, the dilemma remains. At the edge of existing markets, there are considerable numbers of base stations that are not fundamentally profitable. This is particularly true of some of them that require two-generator fuelling and satellite backhaul. You can share infrastructure with operators and rent towers but that only takes you so far. And beyond the edge of existing markets are another set of edge markets that make the current ones look like a walk in the park: low population densities; off-grid and off-network; small numbers of data users (missions, resorts, etc) and low numbers of voice users with marginal and sometimes vulnerable incomes.
The only way to address these edge markets is to lower both the OPEX and CAPEX costs. For example, the Open BTS project tested a network at the Burning Man Festival in the USA back in the summer. It was not interconnected to the national networks but successfully provided a phone service for 40,000 people.
OpenBTS is built on Linux and distributed via a AGPLv3 license. When used with a software-defined radio such as the Universal Software Radio Peripheral (USRP), it presents a GSM air interface (“Um”) to any standard GSM cell phone, with no modification whatsoever required of the phone.
Two of OpenBTS’s three founders are a duo of wireless design gurus that make up Kestrel Signal Processing: David Burgess and Harvind Samra. The third is industry luminary Glenn Edens, the same Edens who founded Grid Systems, maker of the first laptop in the early ‘80s, who is also known as the former director of Sun Microsystem’s Laboratories (among his other credentials). He is the CEO of Range Networks who will commercialise the offering.
It requires a mere 50 watts of power so it is easily supported by solar or wind power, or batteries. It performs as well as any other GSM base station which has a maximum range of 35 kilometers and a typical range of 20 kilometers, depending on geography, antennae height, etc. A full?power base station with software costs around $10,000. Compare that to the typical $50,000 – $100,000 investment for base station controllers, mobile switching centers and “a whole lot of plumbing” to bring in power, backhaul, etc., in a traditional cellular network.
The continent’s own low cost challenger is the Village Telco that is being promoted by the Shuttleworth Foundation’s Steve Song. It uses a Wi-Fi Mesh technology that is low-cost and easy to roll-out and is called with tongue-in-cheek chutzpah, the Mesh Potato. It launched the device at the beginning of September and it is currently being rolled out in Orange Farm in South Africa: it was possible to roll-out nearly 40 devices in just two days.
So the tactical choice for mobile operators is: do we do this ourselves or let some other brave soul throw themselves off the cliff, clutching the latest, all new, low-cost technology? At least one major mobile operator has been trying out Wi-Fi mesh technology but in Eastern Europe, not an emerging market. It seems to me that both regulators and mobile operators should encourage others to roll-out in this way. For as the Godfather said in the movie:”Keep your friends close and your enemies even closer.”
For maybe one day, large parts of the network will be run using this kind of equipment and the kind of assumptions currently made about mobile voice and data networks will change completely.
Mobile industry in Africa << comments and views