Speed is the hidden factor behind the slow-down in growth of online retail in South Africa, writes Arthur Goldstuck. According to latest World Wide Worx report, "The Goldstuck Report: Online Retail in South Africa 2004", the number of retail web sites has grown from 215 at the end of 2001 to more than 700 at the end of 2003. But the growth in online sales has been nothing as spectacular. Indeed, it is slowing down. Online sales by retailers - which exclude property, cars and travel - increased by 35% in 2003, but are expected to grow by only 25% in 2004. A total of R341-million in online retail sales was achieved in 2003 - a mere 0.14% of the overall retail market in South Africa.

Moreover, the market is dominated by the top eight online retailers, who between them account for about 80% of all online retail sales in South Africa. The dominant online retailers are Pick 'n' Pay Home Shopping,, Woolworths Inthebag, Netflorist, Cybercellar and, with the MWEB ShopZone dominating the market in online shopping malls, followed by Digital Mall.

Growth for these players is steady, but not as strong as in past years, when an annual doubling of turnover could be expected. However, most of them are healthier than at any time in the past, and most are generally profitable on an operational level.

The report includes in depth interviews with nine key executives in the e-commerce arena, and reveals the strategies behind the success stories in South African online retail. Most of these pinpoint lack of access to decent bandwidth at decent prices as the single biggest obstacle to growth.

There are suggestions among telecoms providers that South Africans can now get all the bandwidth they need, but this is a rather disingenuous standpoint. Firstly, the highest bandwidth that is close to affordable for existing Internet users is 512Mb per second, supplied by both Telkom in its ADSL offering and Sentech in its MyWireless service. But both are in fact expensive - starting at R680 excluding VAT, and that VAT charge is a real cost incurred by consumers, as opposed to the input cost that businesses can claim back from SA Revenue Services.

This is one of the reasons behind the slower than expected uptake of ADSL, which is hugely popular among small business but a tough budgeting decision for consumers. Another factor is that neither ADSL nor MyWireless are as widely available as their marketing messages would suggest. The former, because only certain exchanges support ADSL, and at a certain distance from those exchanges; the latter, because the roll-out by Sentech has to occur on a site by site basis, meaning that it is some years from general national coverage.

Cost and access are the main factors, but there are others, such as the matter of speed. The reality is that neither of these options represents true broadband. According to Broadband Daily, while the Federal Communications Commission says that connections with speeds exceeding 200 kbps in at least one direction are broadband connections, this definition is not accepted by the more technically-minded.

"Some hard-nosed high-speed junkies say that broadband begins at 1.5 Mbps.  Some policy advocates, who mostly promote a technology overhaul that benefits tech suppliers, say broadband doesn't begin until speeds reach 10 Mbps, and is ideal at 100 Mbps," says Broadband Daily. "The truth is, broadband is relative. For the dial-up user, 200 kbps is broadband, or at least broader band. For the mobile phone user, 200 kbps over a handset is super-fast, at least for now. Cable keeps setting the gold-standard for broadband speed, with most cable companies upping the throughput rates to 2 Mbps, 3 Mbps and even 5 Mbps to keep pace with cheaper, but still-slower DSL competition. Phone companies vow to try to keep pace with cable, but 1.5 Mbps is the prevailing rate for DSL."

The maximum theoretical speed for ADSL is 8Mbps, with 2Mbps the maximum commercially available in a country like the United Kingdom.

This puts our ADSL service's "broadband" in neat perspective: DSL services in the United States are offered at three times the speed of Telkom's ADSL precisely because it is important for the service to be seen as "true broadband" and to be compared to equivalent offerings. Similar access speeds in South Africa require heavy investment in leased lines, at a cost that is accessible only to corporates.

Can we blame Telkom for wanting to make huge profits on phone as well as data services? Of course not. If the telecommunications authorities say that is what they are allowed to do, that is what they will do. But it is time for the telecommunications authorities to understand the damage they are doing to competitiveness in South Africa by not acting more decisively on deregulating the market more effectively.

Online retail may be a tiny market, insignificant economically. But it is strategically important, and it is a benchmark of the performance inhibition that results from the telecommunications framework. A healthy telecoms regime would result in a healthy online retail environment. But it would also advance the health of so many other sectors of the economy, and hence the wealth of the nation.