Telkom to open its exchanges to rivals - TechCentral

Telkom to open its exchanges to rivals

Telkom CEO Sipho Maseko

Telkom CEO Sipho Maseko

Telkom will open its telecommunications exchanges to Internet service providers using an open-access model in an effort to stimulate demand for fixed-line broadband in South Africa.

“To further stimulate access to broadband, we will be making around 200 of our exchanges available to Internet service providers on an open-access basis during 2015,” Telkom CEO Sipho Maseko said in the company’s latest annual report, which was published on Monday.

Though it’s not completely clear yet what the decision means, it appears to be a marked reversal from Telkom’s approach of old, where it kept tight control of its assets, preventing rival service providers from getting access to its last-mile infrastructure.

Indeed, Telkom has for five years fought attempts by rival Neotel to use regulatory processes to get access to key exchanges owned by the company. A Neotel complaint to communications regulator Icasa, seeking access to key Telkom exchanges, is still the subject of a high court application by the fixed-line incumbent.

Telkom is reluctant to provide details about how its plan to open its exchanges to rival Internet service providers will work in practice – the plan is still in the early stages of development, apparently.

Asked for further comment, company spokesman Jacqui O’Sullivan said Telkom intends running “pilot projects based on data uptake to fully realise the profitability of our exchanges”.

“We believe this would be the right basis through which we could create an environment for open access in future,” O’Sullivan said.

It’s widely known that many of Telkom’s exchanges, especially in smaller regional towns, are not profitable for the company.

“We are initially doing this in areas where it is easier to determine total cost of service to refine the model and will look at rolling out open access once we have constructed a repeatable model for this.”

Open access is a model of telecoms that has been pioneered in South Africa by new fibre players such as Dark Fibre Africa, Link Africa and Vumatel. In essence, it’s a model where owners of telecoms infrastructure provide access, at varying levels, to their networks to competing service providers. The degree to which Telkom intends opening its last-mile network is still open to question.

Open access is typically favoured by challenger operators, and not by big incumbents. But in South Africa, MTN and Telkom have now both professed their support of the model, at least to some extent.

Earlier this month, MTN said it now supports the concept of open-access for the deployment of fibre-to-the-home (FTTH) infrastructure in South Africa. It said it believes the model provides customers and independent service providers who use its FTTH network with “flexibility and a seamless and distinct customer experience”.


Tim Parle

Tim Parle, senior telecoms consultant at technology consultancy BMI-TechKnowledge, said Telkom’s plan to open 200 exchanges appears to tie in with its FTTH deployment plans.

The 2015 annual report, he said, reveals that the company intends offering access to its new-generation broadband access technologies – including its multi-service access nodes (MSANs) – to other Internet service providers on an open-access basis.

Telkom is rolling out MSAN cabinets in suburban streets across South Africa to provide high-speed copper VDSL (up to 40Mbit/s) and fibre (up to 100Mbit/s) access to consumers.

“It is not immediately evident what this is,” Parle said of Telkom’s plans, “but my hunch is that these are honeyed words for the wholesale fibre broadband access product. This is to fibre what Telkom’s IP Connect traditionally has been to ADSL and VDSL — the mechanism by which other Internet service providers can access Telkom end users. This is already being used by Vox Telecom, WebAfrica, Afrihost and others in their FTTH products.

“All told, this looks like a progressive move from Telkom, which has realised that it’s not the only game in town. It also resonates with moves in the UK and elsewhere to open up exchanges,” he said.  — © 2015 NewsCentral Media


  1. Greg Mahlknecht on

    Interesting, but we’ll have to wait for full details before getting excited. I think it’s too little, too late. Open exchanges mean that ISPs still have to lease/lay fibre to Telkom exchanges/MSAN to get access to the ancient crappy copper network; it might be better just to go the one step further and roll out the last FTTH mile.

    Bear in mind that a server rack and a tiny room is all that’s needed to deploy FTTH to the houses covered by a few Telkom exchanges – so from the point of view of centralized infrastructure in a next-generation network, Telkom exchanges have little to no strategic value. And in the same FTTH future, the MSANs are simply redundant.

    > “but my hunch is that these are honeyed words for the wholesale fibre broadband access product. This is to fibre what Telkom’s IP Connect traditionally has been to ADSL and VDSL

    That would be my guess too – up to now, Telkom’s (new) FTTH strategy has been to replace the ADSL income with similar prices for similar speeds/traffic, and the IPConnect model ties up with this.

    I believe that Telkom isn’t very relevant in the future of ZA telecoms – they’ve kept ADSL expensive enough to ensure that it’ll be worth for FTTH players to run fibre in those places, and this could be a last gasp effort to get them to delay roll-out or sweat the copper assets before they rust away.

  2. Great strategy from Telkom. Part of Telkom Video on Demand projects?
    Soon we’ll be able to use new spectrum to be released from digital migration as well. Connected TV’s are a growing Phenomenon the world over.

  3. If priced correctly, perhaps co-location at Telkom exchanges could be one of the other strategic value add to be derived by ISPs from this. But the real issue here is whether Telkom will be able and willing to price metro ethernet from these exchanges competitively to enable traffic to be carried to other competitive IXPs like TERACO, JINX. And the issue of the rental for that ADSL or Fibre last mile will remain with us.

  4. Greg Mahlknecht on

    Yeah, but the problem is that the big players in this space like DFA, etc have such massive pipes running along so much real estate, it’s dead easy to plop a server rack in a secure room somewhere and spend R2k on rent. This is what Telkom’s competing with, and it makes their exchanges worth so little, and greatly diminishes the relevance of their Metro Ethernet backbones.

    2 or 3 years ago, it was hard to see South Africa realizing any kind of 1st world internet solution without Telkom – but they’ve done so little since then, it’s now hard to see their place in this same future. They’re still desperately clinging to their ADSL models (which didn’t work for them in the first place!)

    I don’t think that Telkom peer at IXPs yet, do they? Another thing they need to change ASAP.

  5. I cannot see what Telkom’s big idea is exactly as exchanges are by and large old technology now as the planet has moved onto digital years back.
    I also dread to think what the record keeping will devolve into now with so many cooks spoiling the broth at both the MDF and carrier/transmission systems level.
    Apart from easier access to the last mile copper I’m at a loss to see any benefit to anyone actually.

  6. Too little, way way too late. Telkom is only waking up now, because the new-breed fibre operators are stealing their lunch from right under their noses. It now makes very little sense for 3rd party operators and ISPs to invest money into a dying outdated and mostly poor quality copper network, while getting whipped by Telkom in the background. Their money would be better spent in faster, cheaper, highly scalable, more reliable and future-proof fibre-optic networks, that are free from anything related to Telkom.

    This is the single biggest thing I will ever applaud Telkom for. They made the business model for competing fibre-optic networks viable, by pricing it’s own outdated copper products out of the market, not investing enough and not investing quickly enough into its own fibre products, and offering very poor service to consumers, businesses and ISPs. Well done Telkom. Thank you for digging your own grave. You can be sure I will be there to help fill it with gravel.

    As a consumer, I and many others will kick Telkom to the curb as soon as we get access to alternative fixed-line operators. It’s the only way we can ensure that competition and innovation flourishes in the telecoms and broadband market. I will not give Telkom a cent more than I absolutely have to.

  7. Yep, agree : Too little, way way too late.
    Telkom is the main reason of the immensely inadequate and delayed roll-out of fixed line broadband. Only 1 m ADSL subscribers in about 12 years of implementation. On a population of 54.8 M with nearly 20 m regularly accessing the web.

  8. David Robert Lewis on

    Allow ISPs to provide cable, then we will have a big bang of access and affordability. ISPs will then converge into Telcos and we will have a major leap forward economically.

  9. Cable is old school as well, based on copper, and there are NO cable networks in RSA. Fibre-optic transmission is the new dog in the ‘hood. Pointless laying out cable when you can leapfrog it and lay out fibre-optics instead.

  10. this has been allowed since 2009, the suburban fibre deployments seen over the last year are an expression of exactly this.

  11. David Robert Lewis on

    The private projects in gated communities are on the top-end of the market. A policy allowing ISPs to become Telcos and Cable providers would be more efficient.

  12. in January 2009 approximately 543 entities were given licences essentially equivalent to those of Telkom, Vodacom etc (without the spectrum), i.e. at that stage they had the same rights to deploy cable/copper/wireless infrastructure/etc (and they still have these rights). Since 2009 it costs ZAR10 400 to obtain a class electronic communications network service licence from ICASA which entitles you to deploy and operate an electronic communications network in a district or metropolitan municipality (i.e. a lot bigger than a gated community). This was not government policy it was industry forcing the issue through the courts. So this “big bang” has already happened. The restrictions on deploying cable or other networks now lie with wayleave and other permissions and with the ongoing tragedy of spectrum.

  13. David Robert Lewis on

    So are you saying the Munis are the bottleneck problem? I would have thought the solution would be to break-up Telkom into regional telcos, allowing the company to buy wireless providers and vice versa, and then creating incentives for broadband cable. Cable is what creates the efficiencies that lead to the Dot.Com 1.0 and 2.0 booms. Instead we have a wireless sector that is still very much where dial-up used to be and a decade of retarded growth as a result. Dialup was killed by Telkom when it insisted on a per-minute fee structure. If it had been flatrated to begin with, we could have been ahead of the pack, but instead SA, chose to go the path of earlier apartheid regime, which was a late adopter of technology. We only got television in 1976 and broadband cable in 2006.

  14. Greg Mahlknecht on

    > So are you saying the Munis are the bottleneck problem?

    That’s a big part of it, but the main part was just a failure of the private sector – up to recently, nobody’s had the guts to put up the cash and get those trenches dug. But it looks like the floodgates have opened now, and it’s happening at last.

  15. David Robert Lewis on

    All I want is to be able to choose who my cable modem DSL provider is, and ditch Telkom. Why do I have to pay two different companies to be connected to broadband? Its like having a barnacle stuck to my back, or that MTN dog on the leg.

  16. Greg Mahlknecht on

    Indeed. And that’s what’s happening right now. However you’re connected anywhere in the world, you’re paying 2 or 3 companies, they just all roll up in to the ISP bill – but in ZA we have the stupid R160 voice portion you can only pay Telkom for, so we’re stuck with 2 bills.

  17. David Robert Lewis on

    We need naked broadband and choice when it comes to telcos. Telkom monopoly is the reason for retarded growth in South Africa.

  18. Greg Mahlknecht on

    If Telkom do indeed open their exchanges, then naked ADSL won’t be far behind – the problem is because of the way Telkom does their broadband wholesale (both ADSL and FTTH), it’s easy to see from a mile away that the products that result from these open exchanges are going to cost pretty much the same to the end user.

    Agreed 100% that Telkom’s the reason for the retarded growth, because they’ve had CEO after CEO that is useless at their job, and just keeps guiding Telkom down the misguided path of their predecessor, shuffling the deck chairs while it sinks. And this new guy’s not much different.

  19. David Robert Lewis on

    There a number of reasons for Telkom stagnation. Firstly, South Africa needed a break from the previous regime, instead government insisted upon continuity at all costs. Secondly, the Internet as a technology was simply bolted onto pre-existing infrastructure instead of being rolled out separately. Thirdly, at the insistence of unions and the SACP, the state engaged in a dirigiste economy via a state monopoly. All of this retarded growth. We now heading towards a situation where the state is the largest employer and where manufacturing and the private sector is in decline. This hasn’t changed unemployment except by a miserly 1.5 percent, so we treading water. It would be better to introduce competition and keep the state as an investor, instead of maintaining a monopoly at all costs. Monopolies do not equal better jobs. Good to see there is something afoot, but we need bolder plans and better access for all.

  20. I think you are too bullish about fibre. It will take several years before we see anything close to copper’s penetration. There is also a skill shortage in terms of delivering fibre services, which is slowing down the process. Besides, if I can get ADSL at half of what I’m paying now, I’d choose it over fibre. One day I may crave 40mbps+ speeds connecting everything. Right now I’ll be happy if 10mbps costs me less than R400 a month.

  21. From what I’ve been told, fibre is a bit more complicated than that and also requires quite a bit of pre-planning if you aim to include robust services. Were fibre purely a replacement for ADSL, no problem. But given its broader services, that’s creating problems.

  22. Nope, don’t agree. I’ve been told to wait a minimum of 2 years for fibre to reach my suburb – and I’m well inside Jhb. That is the case for many areas in the metropols and pretty much every smaller town and city.

    It will take a number of years before fibre rivals the penetration of the copper network.

    Besides, nothing in fibre services the entry level. There are no packages that compare to the 1-4mbps users, who want cheap Internet. That’s the main growth point for Telkom’s future and since the fibre market has yet to mature, it won’t be able to turn out cheap rival products.

  23. Finally! And this is the first time Telkom actually made a positive move towards LLU.

    Many here say fibre will kill this, but I think this is because of fibre. Fibre is not cheap and nothing in the market services entry-level consumers. Copper is the one thing Telkom has that it can leverage very easily. It’s also mostly paid for itself, unlike any of the fibre networks, which means Telkom can make a low-cost attack on the market. I think it is opening some exchanges in the long-term view to offer cheap Telkom ISP packages without being accused of monopolistic actions.

    But this will mean a drop in profits. The real question is if Telkom’s management has the guts to ride this out. If they drop prices and aggressively pursue ADSL expansion, they can come out shining. But if they keep putting out an inferior product at the same price as its faster rival, they will fail.

  24. Greg Mahlknecht on

    > It will take several years before we see anything close to copper’s penetration

    Indeed, but bear in mind that fibre is being deployed in areas where copper has high penetration; I’m guessing it’ll be close to a 1:1 conversion, so they’ll maybe meet around the 600,000 user mark. ADSL is still a luxury product, like FTTH.

    >There is also a skill shortage in terms of delivering fibre services

    I don’t believe this is true. The manual side of things is just digging, and a little light building; no shortage of that in this country – and the networking skills can easily be imported if not sourced locally. From the user side of things, you just unplug ADSL from the phone line and plug your FTTH router in to the fiber. Nothing special.

    >I’ll be happy if 10mbps costs me less than R400 a month.

    Right now uncapped 10mbit FTTH in some open access networks already R600. I don’t see ADSL being cheaper than FTTH over the long term; Telkom has such crippling overheads they simply can’t be as aggressive with pricing as their competitors. And it’s not that FTTH is particularly cheap – ADSL is just massively overpriced.

  25. Greg Mahlknecht on

    >There are no packages that compare to the 1-4mbps users, who want cheap Internet.

    Not true. Go do some research on this – check out WebAfrica’s site, they have FTTH packages starting at R199/mo, and R350 for a decent 200GB/4Mbit one. Bear in mind you can port your landline number and do away with your R160/mo Telkom voice line in addition to your ADSL package and suddenly ADSL doesn’t look so cheap anymore.

    >It will take a number of years before fibre rivals the penetration of the copper network.

    Agreed here… the roll-out will only get faster as competition increases. Yes, it’s going to be years until we reach ADSL line numbers, but time flies but the writing is on the wall.

  26. Also on Webafrica: 200GB for R149 – capped ADSL. The R199 fibre package is for 20gb. The closest in ADSL terms of Webafrica is the 15GB anytime for R79. There is no comparison.

  27. If Telkom plays this strategy towards cheaper ADSL, your assertion that Fibre will kill it is far too premature. Fibre is not Telkom’s problem. Its high prices are. If this move indicates a much more aggressive pricing policy, then copper ADSL still has a long life ahead of it, fibre nonwithstanding.

    As for overheads, everything can be negotiated. It’s all about the willingness of Telkom’s leadership.

    Regarding skills, there is nothing cheap about importing skills. Nothing cheap at all. ICT Skills are one of the biggest concerns for companies in SA and they are already hiring in overseas talent. Importing fibre skills will only retard its expansion until a proper local pool has been established.

  28. Greg Mahlknecht on

    Do you think Telkom can actually afford to drop ADSL prices much more? According to them, they’re already making a loss on their copper lines. I agree that they have no choice, but really their options are limited here.

  29. Greg Mahlknecht on

    > Also on Webafrica: 200GB for R149

    That’s without a line. With a 2mbit ADSL line it goes to R299. Add your required R160 voice line rental, and you’re looking at R460 for the ADSL you quoted. This is significantly more expensive than the 4mbit/200GB FTTH package @ R350.

    Please try and understand the Telkom pricing before you do comparisons.

    >15GB anytime for R79

    + line + ADSL = R160 + R150 = R389. You’re right, no comparison, just not like YOU think! 🙂

    I have found that the majority of people I talk about FTTH to, like you, don’t actually know what they’re paying Telkom, so FTTH is a harder sell right now – but once they get educated, FTTH demand will increase and deployment will accelerate.

  30. Yes, you are right. I forgot about line rental. So it does compare well.

    I still think availability is an issue. Fibre may compare in cost, but if we are to get the next million connected, we’re going to need copper. All fibre does at the moment is convert existing users.

  31. As I said: everything can be negotiated. Telkom can cut its workforce, it can re-negotiate with its creditors, it can make some big decisions about restructuring itself. These are all feasible. The only time a company is so indebted that it can’t move much is when it enters bankruptcy or is being sold off in pieces. The real question is if the leadership can do that and if the investors will allow it.

    Re: skills – Yes, they can be. Those billions have to eventually capitalise. You can’t just pour money into this; you need a sustainable growth strategy. Importing skills will cause major cost overheads that eats into that. We need many people with local fibre skillsets to make sure labour does not consume fibre profits. So either there will be slow growth or fibre companies will take much longer to realise profits. Which do you think will happen? In my experience, companies are rarely altruistic… They’ll favour slow growth, except where clients are willing to fund the imported skills – aka. big business and public sector.

  32. Greg Mahlknecht on

    >Fibre may compare in cost, but if we are to get the next million connected, we’re going to need copper.

    Couldn’t agree more here – because Telkom has 2mil of the lower income people connected with voice but no ADSL. HOWEVER, the big problem is that traditionally, the higher cost products subsidize the lower cost. Telkom’s market share is going to be slaughtered here and if they are forced to cut prices, they’re going to be even less able to roll out the next million. Personally I think the government should find a billion rand and subsidize those connections – I’d consider it tax money well spent – but that won’t happen. And if Telkom hasn’t been able to connect them in the last 20 years, there’s no way on earth they’re going to be able to do it under more difficult conditions. The short, and unfortunate, answer is those next million are going to be screwed over.

    The absolute best outcome would be if government stopped this slow decline of Telkom, and sold it off to cut their losses. If it could be snapped up on the cheap by someone, they could afford to connect the next million.

  33. >> “…nothing in the market services entry-level consumer.”

    Don’t forget about 3G/LTE and other wireless services. These packages service the lower end of the market quite well. And for lower usage patterns, which is what the low-end of the market would be looking at, it comes out cheaper or around the same price as entry level ADSL. 3G pricing starts at around R58 per GB with Afrihost, and goes up to R497 for 10GB on a month-to-month account, no contract. If you shop around, you can find similarly-priced prepaid deals, promos and even contracts from the cellular operators themselves.

  34. This LLU move by Telkom is too little too late. The name of the game is Fibre Backhaul and 4G coupled with its evolution to 5G mobile technology. 5G is the Next Generation Mobile access technology delivering speeds of 10 Gbps with low latency in the millimetre wave spectrum i.e. 30 to 300GHz. The so-called ITU-R IMT2020.

  35. The debate about FTTH should not close our minds to 4G Mobile access network which can deliver up to 100Mbps. Talk about digging up trenches will be for access will be moot, let alone it being a too costly. This to be followed by 5G (ITU-R IMT2020) delivering speeds of up to 10Gbps for the realization of the Internet of Things. Remember there are more devices to be connected than the total world human population. The core network or backhaul will obviously be DWDM Fibre Technology.

  36. Greg Mahlknecht on

    >3G pricing starts at around R58 per GB with Afrihost, and goes up to R497 for 10GB

    Meanwhile for the same price people living in the suburbs will be getting uncapped 10mbit FTTH services for that price. The lower end of the market really gets screwed, and I don’t see any end in sight for them 🙁

  37. The devil will be in the details. (which of course they’re going to be evasive about for a while…)

  38. Andrew Fraser on

    That would of course require that there is spectrum released. And that someone wants to set up the required towers, everything has a cost.

    Also remember that, unlike Fibre and copper, wireless solutions are necessarily contended in the last mile. So the theoretical maximum of 100Mpps of 4G isn’t often attained due to shortage of spectrum.

    Both wired and wireless solutions have their place. For areas with lower population density wireless solutions are a good fit. For more urban areas, wired solutions seem better – more stable, no contention in last mile.

  39. I agree that additional spectrum would have to released. 4G speeds of 100Mbps downlink and 50 Mbps uplink. would need a spectrum of 20 MHz with full mobility of up to 350km/h and speeds of up to 1Gbps stationary. The 4G system is an end-to-end All IP Digital Network that will be IPv6 based, which supports seamless mobility between different access networks i.e. facilitate not only horizontal handover but Vertical Handover. The resource sharing among the various access networks will smooth the problem related to spectrum limitation and relative to 3G.

    It will seamlessly integrate multiple access technologies and end-user devices ( a combination of network heterogeneity and terminal heterogeneity), anywhere (360 degrees coverage), anytime (24/7), any device services (Device multimodality and reconfigurability) to anybody ( business & consumer market segments). MIMO technology will be used for downlink).

    The terminals will be able to pick the preffered access technology e.g. ad-hoc, wired, WLAN, DAB/DVB, or cellular and move between technologies seamlessly at any location

    The 4g wireless technology will replace the current technology with a single universal technology based on IP with an Internet Multimedia Core network Subsytem (IMS) for the implementation of multimedia services and rich applications.

© 2009 – 2019 NewsCentral Media