Hackett’s premise flawed, says NBN Co

47

NBN Co has defended its pricing approach in the face of strident criticism from an ISP which will be one of its largest customers, claiming comments by Internode managing director Simon Hackett this week didn’t reflect the “reality” of how it would sell services to the telecommunications market after its network was rolled out.

At the Communications Day Summit in Sydney this week, Hackett — who represents the fifth largest ISP in Australia — described NBN Co’s pricing model as “insane” for small internet service providers, warning that none would survive their walk through the “valley of death” transition from the current copper network to the fibre future envisioned by the Federal Government.

According to Hackett, the NBN’s pricing model will only be feasible for ISPs with larger than 250,000 customers – which only five retail ISPs in Australia can boast – Telstra, Optus, TPG, iiNet and Internode itself, which will just reach the cut-off mark.

NBN Co has not responded to a request for comment in reaction to Hackett’s comments. However, in a bylined article published on Business Spectator this morning, NBN Co head of product development Jim Hassell (pictured) argued that Hackett’s comments were based on the flawed assumption that small retail ISPs would want to provide a national broadband service.

“In reality, we think there are other options and likely approaches for smaller RSPs,” said Hassell. “They can take a more regional approach — and indeed, many smaller RSPs have grown much of their business from a local presence in a state or area.”

Hassell added that NBN Co anticipated the development of “a wholesale aggregator market”.

Optus this week confirmed it was planning to offer wholesale services to other ISPs on the back of the National Broadband Network, and it is possible that Telstra’s wholesale arm could do the same. both companies already provide wholesale services to many ISPs around Australia off the back of their existing networks.

To resolve the problem, Hackett has previously advocated a model where NBN Co would provide as little as 14 points of interconnect – which he said would serve smaller ISPs much better. A model with more points of interconnect would serve larger ISPs like Telstra and Optus, he said, which already had infrastructure around Australia.

In addition, the ISP chief wants the cost of what NBN calls its Connectivity Virtual Circuit service providing data to fibre broadband users to come down.

Image credit: NBN Co

47 COMMENTS

  1. Although I’m pro-NBN, this is the very real problem of having the government so closely involved with the market: they are using their pricing model to determine how the market should operate.

    • I disagree – as Hassell points out, the market will choose to operate how it wishes, whether that be on a national scale, or local or regional levels. NBN Co just provides the tools.

      • Yes, but as Simon points out, it is only providing certain tools, which will only be able to be used by certain people. It’s not an open market when you are still forced to buy wholesale services from Telstra or Optus.

        • Again – disagree. NBN Co is charged by the government to provide Layer 2 carriage only – (there will be some service with Layer 3 awareness, granted). The Telstra’s and Optus’s (?) of this world will take the raw carriage from NBN, and add other features and services onto it at the Layer 3 level, and sell it to their customers.

          We have to start thinking the bulk haulage of bits is becoming a utility, much like the bulk haulage of water throw the water pipes, electricity through the power lines, and sewerage through the drains.

          This is evolution. Progress. Redrawing the boundaries – call it what you want.

          • Hi Michael,

            Let me quote you from September 2010, if I may:

            “each POI will be connected together in a two-way backbone ring, allowing each and every provider access to the entire national infrastructure through as little as a single POI. Bigger players will join via more than one POI for the purposes of capacity and redundancy, but to say that every ISP will need to have backhaul to every single POI is rubbish – the network itself IS the backhaul.”

            http://delimiter.com.au/2010/09/08/god-help-australia-says-linton-on-nbn/#comment-18659

          • Let me draw you a diagram (for you and your other personalities):

            http://www.flickr.com/photos/mwyres/5578584016/

            – red circle = logical two-way backbone fibre ring, core NBN network.
            – blue dots = Points of Interconnect
            – pink box = ISP
            – green line = ISP connection to a POI allowing access to the core network

            QUOTE: “each POI will be connected together in a two-way backbone ring, allowing each and every provider access to the entire national infrastructure through as little as a single POI

            Yup – connecting via a single POI gives access to the entire network – (ie: all joined together)

            QUOTE: “Bigger players will join via more than one POI for the purposes of capacity and redundancy, but to say that every ISP will need to have backhaul to every single POI is rubbish – the network itself IS the backhaul.

            Sure. Draw two more green lines from the pink box to different blue dots.

            I’m not sure what you’re trying to say.

          • Holy… but… what the…???

            This is such a beautifully simple diagram and your explanation is so concise. Why doesn’t NBNCo advertise the network and explain it using these kinds of pictures?

            Thankyou for the picture btw. I have a significantly better grasp of how this network is going to work now.

      • Ben is correct, the pricing model that NBN is proposing if molding the market as they see fit in order to fund NBN

  2. A more regional approach? What does the first “N” in NBN stand for again?

    It seems to me that if anyone is making flawed assumptions here, it’s the person who thinks a startup ISP operating on a government-provided nation-wide fibre rollout will want to limit their scope to a single region.

    Implying there’s room in the modem for a “wholesale aggregator market” seems to me to further back Simon’s point: that only the big boys will get to play in the NBN space and everyone else will get a different set of rules – rules that will be set by the same players we’ve already got.

  3. Simon Hacket is 100% correct. The ACCC has mandated an artificially large number of POIs because that is what the larger ISPs (Optus and TPG mainly) demanded. They didn’t want their existing national network to go to waste, and the ACCC could see their argument about “stranding” assets had merit. But it would be better for the Government to pay compensation for those assets, then move to a model with a smaller number of POIs.

    The large number of POIs is pointless. It just means that every ISP connecting to the NBN will have to duplicate fibre between those points of interconnect and their own equipment. The NBN will have plenty of capacity, and there is very little performance to be gained from the additional POIs, so it is a situation where a sub-optimal choice was made just for corporate-political reasons.

    Now, smaller ISPs have to run fibre to every corner of the country, while the NBN has spare capacity there that could be used with no worries. It makes no sense at all. The NBN knows this, and argued against the ACCC’s decision, but now their hand has been forced, so they are becoming apologists for it.

    • A smaller ISP should start at a more reasonable level, and not bite off more than they can chew – for a quick example, you can serve all of Tasmania, NT and ACT from just 5 POI’s – all of which will have decent back haul within no time – that’s one million people, a average of 220,000 per POI.

      • But the /entire point/ of the NBN is that it is a nationwide network that is bringing access to everyone. I think it is completely reasonable for a small ISP to want to start at a national level – as an end-user, I certainly don’t want to hear about a new small start-up in Tasmania that has an awesome service, but it’s not available to me in Brisbane.

        The entire point of the Internet is that it is accessible to everyone, everywhere – you’re not selling shoes from a corner store and trying to grow it into a major international footwear label or something.

        If you go through the motions of starting an ISP (setting up customer management, email services, etc) and you have a nation-wide network to access, those services will be 100% immediately accessible to anyone else that is plugged into that network. There’s no reason those customers should be artificially or magically out-of-bounds to you.

        Unless you don’t really have access to a nation-wide network. In which case it seems we’re just closing the door on proper competition, and it’ll be the same old players making the rules with the same old creepy peering arrangements between them to artificially hike the cost of domestic transit.

        • “you’re not selling shoes from a corner store”

          And why would someone want to start selling shoes from a corner store instead of starting up a chain of national shoe stores? That’s right, it the amount of capital required to do so.

          Is really that unreasonable to expect that any company that wants to have a national presence have a decent bankroll to support that roll-out?

          Your shoe store can set up in ANY/ALL major shopping centre(s), just like an ISP can connect to ANY/ALL POI.

          How many major companies can you name that started off with a massive nationally roll-out?

          • Let’s try a different example.
            You are trying to setup a shoe store with niche shoes.

            Instead of a corner store (bad example) you setup an internet retailer.
            In this example NBN is like your shipping company.

            The government has just committed to building the biggest shipping network in Australia because TNT and Australia post have been neglecting their own. Shipping routes.

            Why would you then price it so only Australia post can afford to send everywhere, and your small internet store can only afford to send shoes to a single region. Or only 5 regions.

            This is the Internet. When you have a small number of customers, you don’t want to limit where those customers come from, you want the largest possible customer base to grab that small percentage (which are evenly distributed) that see value in your model.

            If only 1% of customers see value in your product, chances are only 1 POI isn’t going to give you enough customers to grow. Especially if you have to have a razor thin margin to compete with the big boys that already have enough customers, and don’t need to save money for growth. (like you do, because you need to save to afford future POI’s).

            Overall, it is academic, there will always be a size of ISP that wont be able to afford EVERY POI, but it isn’t as simple as “you need the capital” when effectively we all thought the capital requirement was one we thought the NBN was going to solve.

  4. It’s unfortunate that the ACCC made it harder for ISP’s to sell a national broadband service, but even as is no ISP other than Telstra offers a national broadband service. iiNet who are now #2 started in a garage. There is simply no reason for a ISP to expect to offer a nation wide service from the get-go.

    The ACCC have made a decision and we can’t blame NBNco for it.

  5. My dear Trog, the NBN is a national WHOLESALE network. A retail network can be as local or widespread as it wants. That’s the freedom engendered by decoupling the two, along with uniform wholesale pricing. Despite what Mr Hackett frets about, there will be plenty of opportunities for small and nimble providers to make pots of money.

    • Right, but we’re talking about the difference in wholesale availability which – as I understand it – is the point of Hackett’s complaint.

      Further, it’s not necessarily about making pots of money – it’s also about limiting the scope of available business models. If small providers simply can’t compete or participate on a national scale then it reduces the opportunities available to them.

      • Yes, but Hackett is suggesting that these small ISPs are necessarily going to need to provide backhaul to all 122 POI locations, and by no means is that necessary.

        • Isn’t it the CVC costs on each POI that add most of the cost he is complaining about? I didn’t think he actually calculated the backhaul costs in his first post on this topic, and it was his first post that he declared the 250,000 minimum customers number…

  6. That certainly may be the case; like everyone else I’ll have to wait and see what the network /actually/ ends up like (as opposed to what they’re /saying/ it will end up like), but for now if Hackett is raising concerns I’ll certainly be paying attention to them!

  7. Its worth understanding that in this context, the current NBNCo pricing model plus the ACCC POI decision means that the definition of ‘small’ is ‘any ISP with less than 250,000 customers’.

    There are a number of national ISPs with less than 250,000 customers, and as they are already national, naturally they need, and they want, to be national on the NBN as well.

    The current pricing model so severely penalises such ISPs with high overheads that they’ll be forced to buy via Telstra or Optus wholesale. The irony here is pretty clear: The NBN was supposed to be about replacing the ‘bad’ Telstra access monopoly those providers must use today with a ‘good’ monopoly where every provider could have ‘open and equivalent’ access to the NBN directly.

    The current NBNCo and ACCC cost model means that while ‘all providers are equal’, providers at or above 250,000 customers are more equal than others.

    Put simply – thats just not fair.

    The NBNCo cost model can be re-balanced very easily (lower the CVC price from $20,000 per gigabit per month down to $2000 per gigabit per month, and add about $1.50 per month to the port access price).

    That generates exactly the same income for NBNCo at 250,000 customers and above, but it also lets providers smaller than that choose to be nationally available on the NBN without going broke with massive fixed costs that they can’t otherwise overcome.

    It seems reasonable to expect it to be possible for a national ISP with 10,000-20,000 customers to be able to rationally make ends meet directly into the NBN, but as it stands today, such an ISP will be the permanent economic prisoner of Telstra Wholesale.

    Surely the point here wasn’t to entrench the Telstra monopoly over other retailers – but instead, if not addressed, those smaller retailers will be having to buy from their largest retail competitor in the NBN context.

    This is not the future we were promised. Industry -or- consumers. And thats the point.

    • Thanks for the input Simon – I know you’re a reader here from time to time, and I hoped you would.

      I can see exactly where you’re coming from, and you’re absolutely right that the balance could be shifted on the pricing model, and deliver the same or almost the same pricing outcomes for NBN Co. I’ve juggled a lot of numbers around in spreadsheets, and know that you’re right about that.

      I think we you and I are going to differ is a point of perspective. As the boss of one of the major ISPs in this country, you’re absolutely on the money to be chasing the best possible deal for you organisation. After all, that is your job.

      Internode’s dealings with Telstra in the last 12 months are well documented, and having been on the receiving end of similar “less fair than others” deals with Telstra Wholesale, I understand completely.

      While you perspective is from the higher end of the market, mine is from the lower, and I do honestly believe NBN Co have this about right – no doubt from your comments, you will disagree.

      I do believe that the current model will encourage the small, regional/local ISPs to improve and grow, and that can only be good for retail competition.

      Currently the top half-dozen ISPs probably have about 50% (or more) of the market between them. I don’t see that changing, even under the NBN. The cream will continue to rise to the surface. Internode will continue to be amongst that.

      But if the smaller guys have a better ability to grow, it’s better for retail competition that instead of five or six ISPs dominating the market, if that number doubles or triples, all the better.

      Every ISP will feel some pain adjusting to this – big or small – but overall, I think it’s a better market outcome. I have great respect for you and what Internode delivers, and admire your passion for doing the absolute best for the company.

      Overall, I think the NBN is designed primarily to improve the market for consumers to help bring the lesser players up. They’ll need the right business plan to do it, and some will pull it off, and some won’t. In a situation where smaller players are given incentive to grow, it’s always going to hurt the big guys to some extent.

      That’s where you’re coming from, and kudos to you.

      As I said before, the cream will always rise to the top. The real winners in the “new” market will be the ISPs who can deliver the best value-added services on top of the basic data services.

      Guys like iiNet, Optus, Primus, AAPT, TPG and Internode will still be in the best position to do that – (unfortunately, so will Telstra!) – given their strength in the market already – something that I am sure you’re proud to be counted amongst.

      • @MichealWyres

        “Currently the top half-dozen ISPs probably have about 50% (or more) of the market between them”

        How about the top two ISP’s have about 50% or more of the retail BB market between them.

        “I don’t see that changing, even under the NBN.”

        Neither do I, especially as the top two are also the top two wireless providers.

      • I’m confused. Did you read simons post?

        He was saying if you juggle the numbers, you end up with large ISPs paying the same, (majority of the market) and small ISPs paying relatively less, for the same access to customers.

        Tell me, what would encourage growth best in a small ISP. Lots of POTENTIAL customers, or a small number of POTENTIAL customers? If a small ISP was limited to a small number of POI for cost resons, how would that “encourage growth”. To me, it looks like it encourages stagnation. It DISCOURAGES growth. It isn’t in the small ISPs interest to open up another POI because the startup costs will be relatively huge. It would encourage small ISPs (presumably running a tight ship to start out anyway) to sit on customers in a small number of POI until they had the financial security to extend to another POI, which might not even turn out anyway, they might pick the wrong POI to expand to, and it isn’t like you can pull out of a POI once you have customers on it, because that would be huge bad press.

        Simons post clearly indicates that you can get very similar financial outcomes for the NBN by changing where the costs come in. IE a customer adds the lion share of costs, NOT a POTENTIAL customer.

        • Peter – yes, the financial result for the NBN is the same.

          The backhaul links from a POI to an ISP’s POP are NOT part of the NBN infrastructure. Simon’s point is that by having more of them, an ISP may require more links.

          So yes, the financial result for the NBN is the same, but for an ISP choosing to interconnect their POP’s with more and more POI’s the financial impost MAY be different. They aren’t buying those links from NBN.

    • “but as it stands today, such an ISP will be the permanent economic prisoner of Telstra Wholesale.”

      You mean Optus,. TW has not made any official announcement that they are going to wholesale NBN, I assume you don’t mind the ‘competitive tension’ you are so fond of is there if someone like iiNet is also a NBN wholesaler?

      “Surely the point here wasn’t to entrench the Telstra monopoly over other retailers ”

      You mean Optus is the monopoly CVC supplier (so far), Telstra has not made any official announcement on NBN wholesaling as yet, if they do then it’s not a Telstra monopoly or even a Optus one anyway is it?

    • Simon, would you be able to comment on the difference between CVC charges on the 14 POI scheme vs the ~121 POI scheme we have right now

      CVC charges are something that NBNCo requires to pay off its network, this means that NBNCo would have required the same amount of CVC’s from RSP’s (regardless of 14 POI or 121 POI). If the same CVC costs we have (right now) were applied to the 14 POI scheme, wouldn’t the total costs be the same, just distributed in a different manner (except for the lower infrastructural costs required, which is a different matter)

      Apart from that, I agree with what you are saying regarding both CVC and AVC charges, unfortunately your suggestions most likely will not happen because of financial (and thus political) reasons. NBNCo needs those CVC charges (as outlined in their business case) in order to pay off their network. You remove CVC charges, and AVC charges will have to increase (and by a lot more then the amount you suggest) in order for the network to pay for itself. Unless you are suggesting NBNCo should be payed with taxpayers money, there will have to be a loser/s, thats the reality of the expense of a 93% FTTH network

      • A RSP needs to purchase CVC per POI. This is the whole reason that people are speaking out about it. Simon’s (and others) experience with Tasmania is that a minimum 200Meg is required wen you have 100Meg burstable customers. Do the maths:

        200 (meg) x 122 (POIs) x $20 = $488,000/month EX GST
        200(meg) x 14 (POIs) x $20 = $56,000/month EX GST

        In areas where there is contestable backhaul, one FSA (Fibre Serving Area) is covered by one POI. In rural areas where contestable backhaul is not available, NBN will aggregate (backhaul) multiple FSAs into one POI:

        http://www.nbnco.com.au/wps/wcm/connect/c1364480439fe8a38c19fcc5166da634/logical_reference_architecture.jpg?MOD=AJPERES

        But there will still be 122POIs, so you require 122 sets of CVC.

        Simon’s argument is that since there are now 122 POIs, the NBN doesnt have the cost of backhauling traffic back to 14 centralized POIs from all over the country, so they should be passing that cost saving back to the RSPs, who now have to do this backhauling themselves. ie – the cost of CVC should drop from $20 per Meg to $1 per Meg, because it’s now the RSPs who will be responsible for backhauling traffic all over the Country.

        I think there are a lot of people on these forums unfamiliar with how AGVC actually works at the moment in the industry. This may help to understand why some people are uncomfortable with 122 POIs.

        If you use Telstra Wholesale DSL-L2IG you have minimum 1 ATM or GE interconnect (historically referred to as AGVC) in each capital city, and that gives you coverage for the whole state. Alternately you can buy what is called a long haul AGVC to bring other states back to a central POP, but this is very expensive (cheaper to buy your own inter-state backhaul). Powertel/AAPT have a similar model for their wholesale Layer 2 BBADSL2 service.

        If you buy what Optus Wholesale call RBT (Residential Broadband and Telephony), you only need a single inter-connect, and this gives you national coverage. You can elect to have multiple geographically diverse inter-connects with auto failover.

        So this is what ISPs are used to. It might give you some idea why the thought of moving from a single POIs to 122 POIs is quite daunting.


        • 200 (meg) x 122 (POIs) x $20 = $488,000/month EX GST
          200(meg) x 14 (POIs) x $20 = $56,000/month EX GST

          That doesn’t stack up when you factor in – (and lets use the 250,000 customers argument again) – the fact that with less POIs you have more customers per POI.

          250,000 across 122 = 2049 customers (on average) per POI.
          250,000 across 14 = 17857 customers (on average) per POI.

          You have to provide enough backhaul to support a reasonable contention ratio across all users on each POI you choose to service.

          Also, if you have 2,000 customers behind 1 POI and 10000 behind another, you don’t have to pay for as much CVC for the first POI as you do for the second. It’s completely flexible to allow you to dimension the service on an as-needs basis.

          • “…with less POIs you have more customers per POI.”
            Exactly, this is good!
            The more subscribers per POI, the more aggregation benefits for the RSP, and this ultimately benefits the customer. Conversely:
            More POIs = less subs per POI = less aggregation benefits = bad

  8. Hackett is so used to getting cheap access from telstra that he is upset when he realises the NBN will make him pay realistic prices!

      • Agreed there Simon!

        Derek – Simon and Internode have been having an extended spat with Telstra over pricing in exchanges where Internode don’t have their own gear.

        I worked for a small local ISP some years ago, and we were constantly getting screwed by Telstra Wholesale.

        We’d get really ridiculous pricing.

        We’d often be told that there were “no ports” or “too far from exchange”, then mysteriously Telstra would sign the customers up a few days later and suddenly all was good.

        Telstra – (in it’s vertically integrated days) – has definitely been BAD for the ISP industry – I’m sure Simon could attest to that.

        • Nice diversion, but we are not talking about Telstra here are we, it’s all about the new monopolist the NBN Co.

          • Yes I understand the emotional historical baggage that comes with the two words Telstra and monopoly, and wringing it for all its worth, but in relationship to the discussion subject matter here it is totally irrelevant.

            We are either talking about the NBN Co monopoly, although that seems to be deliberately sidelined in the rush to try and establish the CVC wholesaler as the monopoly ogre in this scheme, which isn’t even Telstra anyway.

        • @MichealWyres

          “Telstra – (in it’s vertically integrated days) – has definitely been BAD for the ISP industry – I’m sure Simon could attest to that”

          Without Telstra there would be no ISP industry, I hear Internode’s resold TW ADSL2+is selling well.

          • Well that is indeed a brave statement, the current ISP business looks quite strong to me, selling a broad range of products from a number of pick and choose wholesalers, with a mix of Naked DSL, ADSL/ADSL2+, PSTN , VoIP, Wireless and IPTV products.

            It also a brave statement in the face of recent comments from two prominent ISP’s about NBN resell and the future viability of medium to small ISP’s.

          • Oh alain, once again it looks like that to YOU, but this isn’t just about YOU… it’s about all Australians… especially those currently being gouged, due to no choice…(again, no not me)!

            Once again to try to refresh that cemented mind of yours, it is the NBN – National Broadband Network… NOT the AABN – alain/advocate broadband network.

            Got it now?

            I know for one who for what ever delusional narcissistic reasoning, believes he is more worthy than anyone else, that may be hard (like the obvious advantages of the NBN) for you to actually comprehend but that is so!

  9. To draw an analogy, should this fictional ISP be able to advertise on channel Ten in Adelaide and at the same price receive air time in all major capitals?

    The advertisers are businesses and so is the NBN, really, why are we surprised?

Comments are closed.