HumanIPO: With the South African government targeting 100 percent broadband penetration by 2020, what are the challenges to implementing such a target successfully?
David Belson: There are a number of potential challenges that can stand in the way of the successful implementation of such a target:
- Definition – depending on how the South African government is defining ‘broadband’ and ’100% penetration’, the goal may be easier or harder to meet. If the target speed is set sufficiently low, it may be easy enough to meet using existing infrastructure and/or without having to make significant additional investment. Similarly, if the 100% penetration figure does not refer to actual speeds/bandwidth, and instead refers to advertised speeds and/or is limited to the availability of broadband connections to specific types of institutions (educational, government, etc.), it may be easier to meet.
- Geography and associated population density – it will be easier to bring broadband connectivity to larger numbers of people in areas of high population density, whether through wired or wireless connectivity. However, in more rural areas, it may be much harder to reach residents with both high speed wired and wireless connectivity – geographic features may make infrastructure deployment extremely hard, and if critical resources like dependable electrical power aren’t available, it again may be hard to deploy the required infrastructure.
- Economic incentives/willingness to invest – in some countries, the government has provided economic incentives, whether through funding, tax breaks, or other vehicles, to spur the deployment of high speed Internet infrastructure. Absent these incentives, network service providers may be reticent to make the investment to deploy new, or upgrade existing, infrastructure, as they can be seen by investors as near-term capital expenditures with an unclear long-term return.
- Affordability – this is related, to some extent, to the ‘definition’ point raised above. If high-speed connections are available to 100% of the population, but are not priced at a level where most can afford to purchase/subscribe, then has 100% penetration truly been achieved?
It is widely suggested that mobile internet is the only future for consumer internet usage, particularly in an African context. What is your view?
It’s not clear that mobile Internet is the *only* future for consumer Internet usage, especially if mobile carriers continue to impose usage caps and overage charges. In an African context, it has been suggested that because there has historically been such limited investment in “wiring the continent” that skipping right to mobile would make sense. This likely continues to be a sensible solution to broadly providing access to broadband connectivity, there will continue to be a place for wired consumer Internet connections as well.
According to your Q2 report, internet speeds in South Africa are slowing. What causes this?
The Q2 report reflected nominal quarter-over-quarter losses for average and average and average peak connection speeds as measured by Akamai. Such quarterly declines are not unexpected, and can be caused by a number of factors. However, in Q3, these metrics both saw quarterly gains. In addition, looking at the graphs that be generated at www.akamai.com/stateoftheinternet, the long-term trends for the average and average peak connection speed metrics in South Africa both show solid growth over the long term.
As more people start to use the internet in South Africa, will this impact on broadband speed? If so, how?
Depending on how infrastructure is/was built out, and how “shared” it is, additional Internet usage could result in lower actual speeds experienced by end users, especially during peak hours, due to congestion, and contention for a limited resource (bandwidth). However, if the infrastructure was built out with sufficient headroom, or is easily upgradeable, or if sharing is limited, then speed impacts may be minimal. (By “sharing”, I am referring to the distribution of the bandwidth to multiple users in a given area – for example, if a gigabit connection is brought to a cabinet, is it shared across 10, 100, or 1000 households in a neighbourhood?)
With mobile operators rolling out LTE services across the country, will this impact on internet/speed in South Africa?
I think that there are a few questions here. Broader rollout of LTE services will likely drive additional Internet usage across the country, assuming that such services, and the required devices, are affordably priced. LTE speeds have historically been rather high, especially as compared to the access speeds available through older mobile technologies/protocols. In some cases/areas, the LTE speeds are also higher than what users can get on wired connections from their telephone or cable companies. In these cases, broader rollout will likely serve to raise the average speeds, aggregated across all connections. However, as noted above regarding “sharing”, as more users adopt LTE-capable devices, limited spectrum will be shared across more users, and the actual speeds that users experience may decline.