Past is past in 10-year ADSL planning

Assuming that the NBN progresses as planned, ISPs need to plan their copper-loop investment strategies as they wind down their ADSL2+ services and shift to the NBN over the next decade. Yet, as a group of providers buys more DSLAMs and then clamours to be reimbursed for benefiting from the competitive market of the past 14 years, it's clear that the reign and importance of ADSL2+ is still not over.
Written by David Braue, Contributor

Many years ago, when you could still get such things, I bought a small CRT TV for my bedroom. It had a great picture, and it served its purpose well — until the dirty, rotten government went ahead and decided to force us all to move to digital TV, with its clearer picture and better channel selection.

This was nastiness of the first order, I tell you, because my analog-only TV is now guaranteed to be nothing more than a big pile of glass, plastic and hazardous heavy metals come 1 January 2014.

I want my money back.

Surely, when government policy forces hard-working Australians and hard-working business owners to have to write off their existing technology and move forward into the future together, we should all be compensated for our lost investments. Come to think of it, in the garage I still have a 2400bps modem that I bought in 1987. I can't plug it into my cable internet service. There are no more bulletin board systems left to call. When I use it to dial iiNet's offices, they just pick up and threaten to call the police if I don't stop prank calling them. Since the government's push from analog to digital networks means that I can't use my modem anymore, I want compensation for that, too.


If we've already gotten enough value out of our past technology investments, can we really complain? (Image by Zaphod, CC BY-SA 3.0)

I am, of course, kidding; I would be more than happy for the government to give me a new TV instead of simply refunding the cash I spent all those years ago. In fact, if we're being compensated for government mind-changing, maybe they should reimburse my mate for the dozen cases of Kev Again In '10 T-shirts that he had made back in May 2010 and still has in his garage.

There I go, kidding again. As, I assume, are ISPs like Internode, iiNet and Adam Internet, who are now apparently pushing for the government to reimburse the millions that they've invested in DSLAMs to build out our nationwide ADSL network. The NBN is going to make all of that investment redundant, their argument goes, and the NBN is the government's fault. Therefore, every ISP that's ever installed a DSLAM should get their money back. Or something like that.

It's a bizarre angle on the evolving and very necessary dialogue about where the local loop will head now that Telstra shareholders have approved shutting it down as part of the NBN Co deal. Assuming the ACCC and Telstra can reach a mutually satisfying conclusion over Telstra's structural separation undertaking (SSU) — and that is, by all accounts, not a certainty, because neither side will yet concede that it's ready to compromise to reach the looming deadline — the industry will then need to chart out a long-term transition plan for the copper access network while the NBN is being rolled out.

During that time, ISPs will be balancing two core internet-delivery services. They'll face the need to keep investing in DSLAMs to service areas where the NBN has not yet reached, whilst playing the cost dynamics of establishing CVCs to interconnect with NBN infrastructure. They'll need to educate customers about the differences, facilitate service migration, and deal with a million inevitable speed bumps that will arise as the transition progresses.

The industry will need to chart out a long-term transition plan for the copper access network while the NBN is being rolled out ...During that time, ISPs will be balancing two core internet-delivery services.

There are costs involved here, but much of this must be done under the NBN anyway. ISPs won't suddenly lose money on new DSLAMs just because they're in transition. If they can't roll out new DSLAMs cost effectively, then they just shouldn't do it; it's a free market, after all. Isn't it?

It seems more than a little bit disingenuous to complain, on the one hand, that the government should pay you for playing by the rules of the telecoms' game all these years, then shovelling more capital into continuing to do so. If an ISP is really concerned that DSLAM investment is so awful, then it should stop installing them altogether and wait until the NBN arrives.

This, of course, isn't happening — if anything, recent announcements suggest that the copper network is poised to get a spit-and-polish the likes of which it hasn't seen in yonks, as ISPs squeeze the last revenues out of the ADSL2+ model before it disappears entirely. Telstra, for one, is embarking on an extensive upgrade program to equip 2000 of its 5000 nodes nationwide with ADSL2+ capabilities, while iiNet and Internode are continuing to pour money into DSLAMs.

Invoking NBN Co's deals with Optus and Telstra to stop offering data services on their HFC networks is a red herring; Optus built its own network, and is shutting it down completely, closing off a revenue stream for them. Ditto for Telstra, which owns the network, but has been bought out of the market.

ISPs buying wholesale transit from Telstra, on the other hand, have not built their own networks, and are, by definition, subject to the changes of their upstream suppliers. If Telstra's copper network is no longer available to connect to, ISPs just need to change to an alternate supplier and keep on going. Conveniently, the government is building one. Most ISPs sell internet access services, not fully fledged infrastructure — this is a fundamental concept of the NBN.

If I run a restaurant, and one day the bakery that I buy bread from goes out of business, can I sue them for cutting me off? Or do I just find myself a new and better bakery to buy from?

ISPs will argue that their capital investment in the DSLAMs is significant, although they don't agree on the amount. iiNet says it spends $100,000 per DSLAM, while Internode's figures are closer to twice that. Either way, the thing is: it doesn't matter how much they put into the DSLAMs. I have it on good authority that the payoff period for a DSLAM is on the order of three years; past that, the investment has paid itself off, and is, like the Telstra and Optus networks, simply a mechanism for generating revenues.

Saying they should be compensated for past purchases of equipment that has now paid itself off is like me demanding compensation for buying the ADSL modem that was no longer needed when I switched to cable. After a certain period, the equipment has delivered its value and takes on a much different financial role.

That said, existing DSLAMs still have another decade or so of revenue-generating activity before they'll need to be shut down. The real questions to ask around this are: for how long will the copper network continue to be upgraded before it's left to coast to its eventual demise? And to what areas will it expand before the case for DSLAM investment no longer makes sense?

Existing DSLAMs still have another decade or so of revenue-generating activity before they'll need to be shut down.

On the first point, it seems safe to assume that investment in new DSLAMs will halt around three years before the NBN is scheduled to arrive in an area. Conveniently, NBN Co has promised to manage a three-year rolling schedule of its roll-out that should help ISPs target their future investments. ISPs privately acknowledge that this schedule will make forward planning much easier; once a suburb appears on that list, these DSLAM economics suggest that it will no longer be a profitable area in which to roll out DSLAMs.

The real challenge is considering how the dynamics of the network will play out in poorly serviced regional areas. It's silly to hope that ISPs will suddenly feel a frisson of charity, and push DSLAMs into marginal areas that they never bothered to service before their investments had the NBN's sword of Damocles hanging over them.

With a set end-of-life in place, ISPs will need to target any future DSLAM investments at areas with enough potential customers to guarantee an ROI — and basic maths suggests that's likely to exclude less-populated outlying and rural areas. Diminishing returns will, over time, progressively challenge the case for DSLAMs in particular geographic areas. Hopefully, NBN Co will recognise this and prioritise those areas for its own roll-out.

The thing to remember here is that ISPs aren't just rolling out DSLAMs for the hell of it; they want to secure as many customers as possible before the disruptive NBN arrives. Winning customer loyalty with new DSLAMs, then transitioning those customers to the NBN, will presumably be easier than prising new ones out of the grasp of competitors — so all this DSLAM investment is actually an ironically timed means to a much more profitable end game. Yet, unlike for Optus and Telstra networks, the game for ADSL2+ is not yet up.

What do you think? Should the government compensate ISPs for killing the ADSL business model? Or will ISPs get their investments back well before the gear is switched off?

Editorial standards