Interview: Alan Knott-Craig

How do you assess the price and quality of mobile relative to other markets?

ALAN KNOTT-CRAIG: If one looks at the operators’ margins on voice and the contributions voice makes to overall revenue, it is clear that tariffs are proportionately high. One year ago, South Africa was ranked 98th in the world in terms of voice tariffs. Things are steadily improving, with tariffs down 34% on average driven by aggressive competition for market share in recent years. Prices are falling not because of lower costs, but due to eroding margins – though this can be compensated by larger volumes. While this trend is bad news for operators, it is positive for the country as quality of life depends on communication and it is a moral obligation to improve access by making it more affordable. A barometer for whether or not prices are too high for people is the degree to which people are porting between networks. The wealthy move networks more slowly as savings for them is less of a pressing issue, whereas the poor change networks on a daily and often hourly basis.

One positive trend for the country is that pre-paid pricing has become on par, and is even cheaper in some instances than post-paid. As prices are lowering quite rapidly, people do not want to get tied into long contracts. And we are having to react by offering post-paid customers greater flexibility to follow good deals as otherwise they will avoid signing up to contracts at all. People are starting to better understand and scrutinise tariff structures whereas in the past they might have felt resigned to accepting them without question.

On the data side, the landing and installation of more and more optical fibre cables internationally and locally are playing a role in lowering the costs of broadband. And data in South Africa is very competitively priced.

PIENAAR: There are innovative forces driving competition in both voice and data. In the past three years, our pre-paid rates have decreased 28%. Price competition in the data market is even more aggressive and over the past two years our price per megabyte has fallen by 60%. It is important to distinguish headline prices, which create a perception of high and sticky retail prices, from the actual prices being paid in the market that are driven by promotions and discounts, and which the majority of South Africans experience. As promotional pricing strategy is also driven by network load, dynamic pricing also contributes to a better overall network experience for customers.

JOOSUB: The effective rate for prepaid is R0.71 ($0.09) per minute, which is very comparable to the rest of the world. On the contract side, our rates appear higher because of the level of subsidy available for handsets in the South African market. One way to address what appears to be a high level of tariffs would be offering customers greater choice when they walk into a retail outlet as to whether or not they want to use that subsidy as a discount. Being transparent that there are separate tariff rates with and without the device, and that they vary depending on the quality of the device could ultimately help resolve the confusion.

We are evolving from a voice to a data-centric world; with growth in voice consumption slowing and data consumption rising. Tariff structures moving forward will be very different with customers being offered a bundle of voice, SMS and data all included in one. Data pricing in South Africa versus other markets is lower and competition continues to drive prices down. And as data evolves to comprise the core of bundle requirements from customers, this means that mobile prices are becoming more attractive.

How do you foresee the adoption curve for long-term evolution (LTE) evolving?

PIENAAR: LTE has the potential to allow the South African broadband market to make significant strides in future connectivity and customer experience. At the same time, spectrum and frequency allocation remain a key challenge. We expect to see device manufacturers producing more handsets that are LTE enabled.

What will remain important for operators is the ability to test and ensure that we only approve LTE handsets that allow customers to fully benefit from the opportunities and experience LTE offers.

JOOSUB: LTE is needed to ensure proper broadband access for the entire country. Data, as with voice before it, should be made available to 100% of the population, whereas at the moment only around 83% of the population is covered. Definitely, the high-end data users will adopt LTE a lot faster. And at the moment the main constraint is the number of LTE compatible devices out there. As LTE devices become more popular, speeds will become four to five times faster, allowing users to access a whole range of new applications. An additional benefit of LTE is that migrating users to additional capacity on the 4G network also offloads some of the pressure on, and improves, the 3G network.

KNOTT-CRAIG: The issue with LTE is not so much about whether or not the technology is advanced and an enhanced user experience, but how much it is actually needed. In the short term, LTE will not have huge penetration and impact. To think the country can simply throw aside the 60-70m handset devices currently in use to be replaced with LTE compatible devices is naĂŻve. For the most part, those who are taking up these advanced LTE services are those for whom money is not an object and want the speed and experience regardless of what it costs. They will become heavy data users, and unless new spectrum is issued out, the network speed will deteriorate, negating the experience they were hoping for. So in all, premature LTE rollout reduces the quality of existing services to provide a service that very few people can actually afford to enjoy.

What would you prescribe as the optimal approach towards distributing access to spectrum?

JOOSUB: Spectrum should be allocated in alignment with the country’s 2020 broadband development strategy, which seeks to ensure access to broadband for all South Africans. The National Planning Commission has laid out of a vision for where the country should go, and making this plan operational requires an enabling framework that instructs different parties on the role they should play towards achieving these objectives. If the country is going to succeed in reaching its targeted economic growth, it is vital that every household has broadband access; industry and government need to cooperate if this is to be achieved.

Broadband universal service is a concept being used successfully in a number of European countries, where the government demarcates where and to whom it wishes to offer universal access, and then puts the contracts to provide the broadband service out to tender. The difference between the usual revenue the winning party would often receive to provide this service, and what they actually receive by government, is subsidised by a universal service fund in which all participants in the industry are required to contribute in order to pay for the underlying infrastructure.

KNOTT-CRAIG: Something to be considered by government is establishing a national network. With spectrum allocated to a national entity comprising a half dozen or more private operators, government could also be a shareholder, but it is key for the private sector to contribute most of the money and to run the network.

While people are calling for spectrum to be urgently allocated, government must be careful to do it right, as once spectrum is issued it is gone. It is difficult to issue it to a company and then try and take it away. If government goes down the wrong road it is very expensive to reverse and turn back, and the taxpayers end up paying twice. Issuing spectrum in bits and pieces will never be as efficient as issuing it bunches. And perhaps, the country should consider doing something similar to what Russia has done, through setting up a national broadband network and forcing competition onto the retail networks.

PIENAAR: The allocation of spectrum is central to the rollout of an LTE network to enable the provision of much needed broadband access at higher speeds. The Independent Communications Authority for South Africa has not made new allocations available to operators and there is currently no clear direction on when there will be additional allocation, and what the qualification criteria to be awarded would be. With that said, we are still rapidly deploying LTE services across the country and believe that spectrum allocation toward operators that are readily able to rollout LTE will best support and propel the government’s vision of broadband for all of South Africa by 2020. We will run out of road very quickly on re-framing existing spectrum and the only viable business case for effective LTE deployment must leverage current mobile and fixed infrastructure of existing mobile and fixed players. Delaying allocation is essentially to delay a better and broader, speedier and lower cost, broadband experience for all South Africans.

In what way can the deployment of broadband in remote areas be ensured?

KNOTT-CRAIG: At the moment, the market is being served with voice, SMS and data merging into buckets of price, speed and quantity. When quality and availability goes up, people want prices to go down by a proportionate factor. Since the early 2000s, countries have allowed the mobile operators, under market forces and competition, to determine and deliver the nation’s mobile strategy. This worked fine, but with LTE the environment is slightly different, especially in a country like South Africa which is lacking the broadband penetration and speed found in Europe and North America. I am not advocating for the government to be the sole builder of the network, as government never does as good as a job at delivering infrastructure as the private sector. In emerging countries, the challenge is not in providing high-income earners with broadband, but in connecting schools and clinics in rural areas with broadband at affordable prices. This will not be achieved by operators alone as it is not financially viable.

PIENAAR: The release of digital dividend one and the expected second digital dividend are well-matched to the provision broadband to underserved and remote communities. These dividends, in conjunction with the licensing of spectrum to mobile operators, should receive significant focus from government in an effort to allow South Africa to achieve its stated ICT policy.

It is anticipated that in two years around 85% of the population will be covered by 3G, delivering speeds of between 1.8 and 42 Mbit/sec, which constitutes a solid broadband experience. Coverage costs rise exponentially as the population becomes less dense. Realising 2020 “Broadband for All” and to reach the remainder of rural areas with 3G, is subject to access to digital dividend spectrum, and will require public and private cooperation, and the smart use of development funds. The investments required are colossal and some areas do not offer a retail-pricing environment that could make it profitable on a standalone basis.

4G is the next frontier and an important piece of economic infrastructure for the country. Spectrum allocation and large capital expenditure requirements are not the only factors that will impact on network quality and reliability. The authorities need to set out “Rapid Deployment Guidelines” specifying how government agencies should cooperate with one another in terms of regulatory approval of site developments and operator planned network investments. The current lack of coordination and prolonged delays in obtaining permits for site acquisitions is a significant bottleneck toward future network planning.

JOOSUB: The government has devised a good national plan and what is now needed is an alignment between industry and government. There are approximately 40m households in the country, and to ensure universal broadband delivery requires specifying those households to be connected through fibre and those via wireless, and developing an ecosystem for this to happen. Duplication is unnecessary and costly. For wireless, as has been discussed throughout, spectrum allocation is clearly needed. A more coordinated approach is also needed toward fibre deployment to minimise road disruptions when laying national long distance cable.