Premises passed the only useful NBN measurement


analysis The National Broadband Network Company and the Federal Government should standardise on the “premises passed” statistic to measure the network’s progress and stop using the confusing and amorphous “premises commenced or completed” measurement to provide concrete detail on how well it is progressing against its network rollout targets.

Last week, Communications Minister Stephen Conroy and Finance Minister Penny Wong published a media release detailing the release of NBN Co’s latest corporate plan covering the years from 2013 to 2015. As it made a number of statements regarding the progress of the NBN in general, Delimiter subsequently invited readers to fact-check the media release, in the interests of keeping the debate over the National Broadband Network objective and based on fact. This article is a follow-up to that article and the discussion amongst readers it created.

On the whole, most readers agreed that the majority of the media release was factual, as it generally attempted to provide a summary of information available in NBN Co’s new corporate plan and didn’t venture too far into controversial statements based on opinion without evidence.

The plan does indeed detail, as Conroy and Wong laid out, the fact that the NBN project is projected to make a return on investment of 7.1 percent for the Government’s investment. It also lays out the possibility of falling wholesale broadband prices, a small percentage increase in the capital cost of the infrastructure build, and a six month delay in its rollout. Furthermore, Conroy and Wong are correct in that the finalisation of a number of contracts, such as NBN Co’s deals with Telstra, Optus and various construction and technology companies will give the project as a whole a greater degree of certainty. There was some minor debate around the issue of wholesale prices, but we don’t consider this to be substantial enough to call Conroy out on it.

However, there was one point which Conroy and Wong made which was debated by readers and which Delimiter believes to be misleading. This refers to the claim that the Government is “on track to meet its target of having work for 758,000 fibre premises commenced or completed by the end of 2012”.

The issue of how the NBN’s progress should be measured has been a controversial one for some time. In February, for instance, Shadow Communications Minister Malcolm Turnbull made the point that NBN Co and the Government had stopped publicly referring to the previous “premises passed” benchmark, and had started referring to a new benchmark, which referred to a figure which included those areas where its network was active, where rollout activity was currently underway, and where work was due to start up until the end of the year.

Using this figure, Turnbull said, represented an “Orwellian change to the NBN definitions”. “The NBN has stopped publishing when people can expect to have internet services switched on at their premises,” he wrote at the time. “The NBN is now measuring how many premises they are working on, planning to work on and are planning to begin planning for. This Orwellian change to the NBN definitions is cold comfort for the millions of households, businesses and consumers who have spent the past five years waiting for their broadband to be upgraded.”

We agree.

Although the “premises passed or under construction” figure used by Conroy and Wong in the media release and also used by NBN Co in its publicity materials gives more of a well-rounded picture to illustrate just how much activity is currently under way with respect to construction of NBN Co’s fibre network, it does not represent a firm figure which shows how many premises NBN Co’s network actually reaches to right now, and it should not be used to judge NBN Co’s rollout progress.

It seems obvious that NBN Co and the Government could easily inflate the “premises passed or under construction” figure up at certain points by commencing very preliminary construction in some areas but not actually completing that construction in a reasonable time frame; meaning that it is not a suitable statistic to measure the NBN’s rollout progress by. If the public is to measure the progress of the NBN by its latest corporate plan, as Conroy baldly invited commentators to do in his media release, that progress should not be measured by using such a nebulous measurement.

In its corporate plan, NBN Co provides two other measurements with which to judge the pace of its rollout. The first is defined as “premises passed or covered”, which is defined as being a state where the “shared network and service elements are installed, accepted, commissioned and ready for service, which then enables an end user to order and purchase a broadband service from their choice of retail service provider”.

The second is defined as “premises with active service”, which is defined as being a state where a valid service order is received to install the dedicated optic fibre cable connection to the premises’ optical network termination device. In short, the first measurement is counted when NBN Co has finished deploying fibre to a premise, and the second when that premise actually signs up to use the service.

Delimiter is strongly of the view that NBN Co, the Government and commentators should only refer to the “premises passed or covered” benchmark to judge the speed of the fibre and wireless network rollouts. The “active service” benchmark is less appropriate because it includes factors outside NBN Co’s control — such as when users decide to actually sign up to the NBN connection already delivered to their premise. This can also be influenced by the marketing activities of retail ISPs. While it is an interesting statistic to gauge early interest in the NBN, it is also fairly irrelevant when you consider that almost all current fixed broadband customers will eventually switch to the NBN’s fibre eventually, as their existing copper and HFC cable connections are switched off by Telstra and Optus.

With this in mind, the following table, taken from its corporate plan, illustrates NBN Co’s premises passed targets for the next few years until 2016. This is the benchmark which Delimiter will be holding NBN Co accountable for, and we recommend other media outlets do the same:

To sum up: Ministers Conroy and Wong: Although such a measurement can provide additional colour, Australians do not fundamentally care about statistics detailing how many premises NBN is currently constructing fibre to. Australians do care about the number of premises which NBN Co has actually connected to its network and which are ready to start using fibre, as this is a hard measurement about actual progress of the NBN rollout and represents an accurate judgement of which areas of Australia are “NBN ready” and which not. Any other figure is too amorphous and open to interpretation to provide a real judgement on how fast the network is being rolled out. We recommend all concerned start referring only to the “premises passed” benchmark to measure the progress of the NBN.

Image credit: Sambo9, NBN Co Corporate Plan


  1. What matters is that its rolling out.. It will reach everybody in reasonable time.. Trouble is everybody wants it now, even the naysayers.. (thats ironic)
    The schedule has only changed minimal. Costs have increased only slightly but mainly due to buying all of Optus cable customers, which helps bring the revenue forecast slightly higher to 7.1% RIO. Speaking of revenue, look at this chart ..–2–620×349.jpg Whats everyone concerned about? They should be really concerned about the NBN being stopped and the huge cost that it will entail and the additional lost opportunity for decades.

    • I hate the NBN, I hate everything about it, noone has any use for it, noone will every have any use for it… and I demand that I get connected to it immediately. Yes, I also see the irony in this position.

    • Speaking of the return of 7.1%, does anyone else find that it is a pretty precise figure – in fact too precise. If you take the NBN to cost $45 billion, that makes 0.1% a maximum 45 million dollars. The project could easily have uncertainty of that amount – if not way more than that.

      • Precision != Accuracy.

        Basically they are indicating they should be slightly more profitable now then before, though what that actual number is could vary. Of course, it would be pretty hard for NBNco *not* to achieve that number given it’s control over pricing and profits. If profits go down it can raise prices, if profits go up [which I think is far more likely] they just drop prices.

      • Geoff: they are able to be very precise for one reason.

        They have a mandate to reduce wholesale costs, but only after achieving their 7.1% return.

        They can make their 7.1% return, the forecasts are showing that clearly, but the amount of reduction in wholesale prices will be determined by how much they exceed the 7.1%. So if they make 14.1% return (in any one year); we will get a 7% reduction in wholesale prices, they WON’T pay the government back sooner.

        I wish they had some leeway to pay off their debt sooner, but either way. The government has decided they will pay off the network in a certain amount of time, and are choosing to invest any excess into the reduction of wholesale prices. Obviously; since they decide what their wholesale price reduction will be, they get to precisely hit their 7.1% return on investment.

        Their mandate is to keep wholesale prices equal or less than current. There is always a slight chance that people will stop using the internet the same way they are now (downloading large volumes) and as such stop using wired internet (to say it is slight is to massively overestimate the chance of this). But in this case; I believe they will reduce the 7.1% repayment amount. (in the same way that they aren’t paying 7.1% right now to the government because they havent got any customers). (this paragraph was brought to you by the “don’t misrepresent what I say and claim they will raise wholesale prices department.”)

  2. I think both premises passed and premises under construction should be used, and quoted separately. To use only completed or completed + under construction doesn’t give the full story.

    • +1

      I agree absolutely. Complex things cannot be adequately captured by single numbers, reporting both numbers separately would give the truest representation of progress.

      That is, while I conditionally agree with Renai that at the end of the day the number of completed constructions is the more important figure, nevertheless the number of constructions underway is also an interesting piece of information. It describes how dynamic, how active, the NBN build process is.

  3. Perhaps “Premises covered by construction contract” would be better.
    Then people would at least know if its going to get built OR Cancelled by the Coalition.

  4. Good article, Renai.
    I agree this is a major problem with NBN reporting.
    Passed/Covered is the most important statistic.
    By all means provide a Commenced value. It’s nice to know, but provide it separately.
    And agree that Active Connections is useful for getting an idea of take-up early on but ultimately of little relevance.

  5. I can see why they want to report the passed + commenced, it gives a stat on how much of the build has been commited to, but its a two edged sword,

    From an NBN perspective, active vs passed gives a real stat of the uptake (% of potential users), while active vs passed + commenced makes it look worse than it actually is.

    If there are 30000 active connections, and 300,000 passed premises, its a flat 10% uptake rate, while if there are an additional 600,000 that have had the build commenced, then the uptake rate drops to 3%…

    I know which number I’d prefer to be looking at,

    Give us ALL the numbers, and let people report how they want. Right now, the numbers provided give a poorer picture than reality.

  6. I agree with the article.
    Maybe I’m hopelessly naive, but I believe labor has shot themselves in the foot (again) by trying to do the right thing.
    They chose differing communities for initial rollout. They chose challenging terrain at times, a challenging customer base. They also chose those communities most likely to benefit from an upgraded internet service.
    They could have rolled out the NBN quite quickly in various CBD’s, most likely using pre-existing underground conduits. Found a tech biased and eager populace, ready to sign up to competitively priced, faster internet connection.

  7. Good call Renai.

    We need a standard reference point if we are going to look at how successful the roll out is and I agree that this is the appropriate one.

    Using anything else wold be a bit like including the weight of the container when you buy 1kg of margarine. You wouldn’t know how much margarine you were really getting.

  8. I agree with the definition of premises passed or covered as being most important for what they did with the money that year.

    Looking at the table in the article, does it seem reasonable that in fy2012, only an aditional 11,000 brownfield premises and 9,000 wireless premises were passed? Seems very low to me.
    It also seems difficult to imagine that if you can only do a total of 20,000 brownfield premises in fy2012, you can ramp it up to complete an additional 400,000 premises in fy2013! Wow that is some ramp up seeing they have been at it for 5 years.

    I hope these figures are correct but get ready for lots of excuses next year when the forecast 400,000 additional brownfield sites gets reduced to 100,000 (my guess).

    • The easiest way to show the progression through these numbers is using the NBN rollout map on the NBN website.
      The tick boxes down the bottom of the page roughly show how the rollout will progress.
      The purple “service available” areas will match roughly to the FY 2012 available sites.
      Given that there is an estimated 12 months timeframe for construction, the FY 2013 additional sites will match up to the “work commenced” orange areas on the map.
      The balance of the pink “work commences within one year” areas on the map will roughly match the FY 2014 additions, with some of the green three year area’s too (a couple of months worth).
      All the areas in the green “work commences within three years” should be included in the FY 2016.

      So the announced rollout plan should match back to the timelines of the passed premisis.

    • Well any project has a starting phase where its mostly preparation work. In the case of the NBN, thats been the last 3 years or so, while other aspects have been worked on.

      Look at other threads to get a better picture (costs on track, etc), but the short story is that while the deals with Telstra were hammered out, not all that much could actually be visably done regarding rollout. They could trial various parts to see what worked and what didnt (which is why parts of Tassie will need to be upgraded sooner rather than later), or get other parts not reliant on 3rd parties sorted (satellites for example), and get progress in those areas, but what people think of in terms of rollout simply had to wait.

      But now the delays surrounding Telstra have been sorted, all the trial areas signed off on, and those various other preparatory things done, then yes, it can proceed full steam ahead. Right NOW we are at that point with the previous 3-5 years all leading up to the next 12 months.

      Ignore whats gone before 1 July, just look at what happens from that date to 30 June next year. So much rollout or commencement in that time, being a massive scaleup from the trial phase.

      Its the next 12 months that will show whether The Plan is a good one or not. All the talk and debate, from both sides, will be proven or disproven between now and July next year.

  9. Oh please tell me that none of the 400,000 additional brownfield premises passed in 2013 come from swapping over the 400,000 brownfield sites that Optus just signed over. Hopefully they are completely unrelated, but I am unsure where they fit into the equation/table.

  10. I’ve always believed it should be reported as ‘premises passed and ready for service’. I know why Conroy has used the other- because it makes it appear like they’re busy.

    While I understand his defense, I think they should drop the ‘premises under construction’ as the official ongoing reported figure. It’ll just lead to confusion and more Coalition FUD.

  11. I agree premises passed and premises connected or ready for service should be separate measures.

  12. Turnbull says the NBN info is ‘Orwellian’.

    Interesting choice of words – perhaps he was thinking of the Coalition’s inability to provide any meaningful detail of what they will do after they abort NBN, other than some real (Shadow) Ministry of Truth waffle about ‘faster and cheaper broadband’.

    Which, like the original Ministry of Truth, is an example of ‘words mean whatever you want them to mean’.

  13. If you’re just trying to measure the progress of the network roll out then, as you say, premises passed and ready for service is the appropriate measure.

    There’s a lot more that needs to be measured to judge the NBNCo’s progress against its corporate plan though. What if the roll out is on track but they’re way over/under on capex or way over/under on revenue? Focusing only on the network roll out is fairly pointless on its own

    I think the NBNCo should report (monthly?) active services by speed tier and technology so we can judge the accuracy of their revenue forecasts. I think they should report capex against plan so we can judge the accuracy of their cost forecasts. You can bet your life they’ll be tracking it. Why not report it?

  14. Great comment.

    you just have to look at the Victoria Park “roll out” in WA which has been on the list of premises “under construction” for over 18 months

  15. I agree, it is much moire accurate to measure progress by what has been accomplished ands not what you are trying to complete.

    Before I would consider using active contruction sites (etc etc) as a statistic it would be revealing to include the average time a premise spends on this statistic before being moved to a completed category. This would allow people to estimate the progress or otherwise a suburb could be in that category for an inordinate amount of time and we would be unable to tell.

  16. Interesting;… there is only 100-150K homes constructed every year so where does the 240-250 thousand homes come into greenfield for 2015 and 2016? (amongst other lies)

    • I don’t think every house is built in a new estate at once! NBN Co on the other hand would be building past every lot in the estate regardless of whats on the property (assuming something will be built there)… not hard to figure that one out

  17. Finally an article from Renai that actually has some common sense. You must enlighten us to what you are on or off this week to finally come to grips with providing such an article. Whatever it is, stick to it.

  18. Sorry Josh but it is hard to “figure out”, on average 150K houses (maximum) built per year in Australia currently, so if home lots are sold and not built on in one year they are built on next year but it doesn’t change the annual count. Also,don’t forget the brownfields is captured in another figure within the NBNCo plan.

    NBNCo have quoted 240K/2015 and 350K/2016 homes connected. This is basically untrue.

    • “on average … (maximum)” probably isn’t a good start.
      But, since your statistic is houses built per year, and there are 6 years listed in the graph, then that is 900,000 houses. NBN Co have forecast to connect 763,000 of them.
      Why do they have to be connected in the same year that they are built?
      And, at first glance the difference is probably the houses being connected under Telstra’s responsibility.

      I’m not sure why brownfields were mentioned at all. Are you suggesting that if a year lapses then greenfields are suddenly considered brownfields?

  19. No the only good figure to use is The “active service” benchmark.

    This an excellent gauge as to how people actually value having extra super duper fast broadband when it not forced upon them by removal of the alternative.

    Even if every agreed in principal the NBN is “good thing” – we see how much value is placed on it when people actually have to put their hands in their pockets. – Money talks – bs walks

    • oops hit send too soon – meant to edit the only uselful metric is the percentage of active services vs connected services

      • I would totally agree , if all people were granted the opportunity to move across when the fibre turns up on their door step. Contracts with Telco’s and ISPs (and certain telcos rushing to sign more people up on 2 year contracts) will hold back the subscription rate… well unless the ISP/Telco can operate for less on the nbn and has an advantage to move you onto the network ;)

      • Whilst the intent of your proposed KPI (essentially take-up rate) is ok, it leaves a bit to be desired:
        1. As has been mentioned, take-up rates have been bought. Showing take-up rates prior to forced cut-over will still provide a useful metric.
        2. In a simplistic scenario, NBN Co changes their entire build to on-demand. Connecting only premises that will take an active connection effectively games your statistic, yet results in a profoundly more expensive build.
        3. It alone does not give any indicator of the rate of rollout. Nor of the financials.

        I would argue that the key factor in measuring the rollout has to be how many houses are connected. As this is where the costs are, it will give you a better picture of whether we are getting value for money. This is based on the assumption that take-up rates are assured.
        Naturally, we should be looking at as many metrics as we sensibly can.

    • Why? People are going to be “forced” onto it anyway, so who cares how many take it up before they’re forced to?

      Clearly, premises connected is a better measure because it’s the number of people who will eventually take up a service (not including the small number of people who don’t have a fixed line connection at all anyway).

      • Given that the government is underwriting this, the fact that a very small percentage are actually using the NBN indicates how highly we desire/want/need it.

        I followed Conroy’s suggestion to look at Whirlpool , and found a thread “what s the killer app” – the OP in that thread said seeing Korea has had super duper fast broadband over there for a few years what is the killer app over there? and given the govt is saying “build it & they will come” – it has been built(in Korea) what has come ?

        There are 300 posts that basically say “gaming”, “fast entertainment downloading, internet TV.”

        Sure there are a few that refer to e-medicine – I rang my doctor on Saturday for an appointment, the earliest available slot was late today – so except for some specialist servicing remote areas (who will be in the 7% anyway) I can’t see that happening either. Some mentioned distance ed – well I did an entire degree online – Adsl256 (all that was available at the time) was sufficient.

        • Well Diogenes, the Vic government (Lib – anti-NBN) recently had Delloites compile a report and they said…

          “Demand for faster broadband services is expected to grow significantly. Almost 90% of households will demand at least second wave broadband, representing substantial growth on current demand. **In 2016, it is projected that 80% of households will want third wave broadband, compared to less than a quarter today.”**

          Also, Phil Ruthven CEO of ISISWorld (iirc) has said, they have apps ready to release here, they just need NBN speeds to do so…

          But you sound like a decent person (unlike most of the political puppets who oppose the NBN because they must). So seriously, you can’t see, having already seen the improvement in comms over the last 10 years, and the ten before that, similar happening over the next ten?

          Regardless, it’s gratifying to see the “taxpayer” funded myth has been busted and many NBN deniers are now admitting it’s underwritten by taxpayers, not taxpayer funded.

  20. There’s only one meaningful measurement, premises connected to the NBN. Stop fudging the figures.

  21. It is a bit pointless to talk about the NeverBuiltNetwork after 2013 as the hopelessly incompetent and corrupt ALP and their Communist Greens controller will be past tense after 2013.

    The Coalition will close down the ineffective bloated nBn bureaucracy as it is repeatedly demonstrating it is useless and riddled with public service style incompetence and corruption and it will be taken over by Telstra who are already constructing key parts of the NeverBuiltNetwork in preparation for this.

    The tiny little bit of the NeverBuiltNetwork that is actually completed by the corrupt incompetent nBn by the election will be simply handed over to Telstra to be then built in a responsible competent manner.

    The “unbreakable water tight nBn contracts” will be canceled by the Coalition just as quickly as the utterly ridiculous Communist Greens “CO2” Socialist TAX will be and the leaky boats will be stopped.

    Stand by for the Coalition to return Australia to PROSPERITY starting next year!!!!!

    • Hey, that’s great bit of satire, sending up those ridiculous political dinosaurs who don’t understand NBN.

      It WAS satire, wasn’t it?

    • @NeverBuiltNetwork…

      Err it is already built in many places, it just isn’t complete…

      Try again.

    • What about this NeverBuiltNetwork?

      The Coalition actually want Labor to win the next election, build the NBN, so that the following election when the Coalition are elected they can then sell it, up the GST to 12 or 15%, sell more of our gold at a ridiculously low price… all to fill the $11B hole found in their previous costings… still leaving them with a wad of cash (just like last time with Telstra) to again say, oh look what good fiscal managers we are.

      A silly politically driven comment – yes… intentionally so to prove a point. Two can play silly politics.

      If anyone actually believes one side of politics is perfect and the other hopeless, they are blind fools, imo.

    • You’re probably right NeverBuiltNetwork…

      Only where are wrong is that it would be hard to absorb a non standard network into existing networks eg. telstra.

      It would probably be ripped out. But note, the majority of cost was not in the fibre build but in setting up the company, that comes at a high price which is intellectual property through paid consultants.

  22. And here’s my folks in a new greenfields estate with a house passed by fibre and all the NBN gear installed but no connection to the NBN, and no timeframe.

    Now they can’t even get ADSL since copper hasn’t been, and won’t be, installed. Satellite is out since they are in the NBN fibre foot print, so they are stuck on 3G for an indefinite period.

    Haphazard, reactive policies that lead to poor planning, implementation and higher costs.

    • @Troy

      May I ask if they’re in an estate with more than 100 lots?

      If not, that was the fault of the government and Telstra not agreeing on Greenfields rollout. NBNCo has only just been handed Greenfields with more than 100 lots. They will be building mobile FSAM’s over the next few months to service all those Greenfields before they actually pass them with transit fibre.

      This build is an entire upheaval of the telecommunications network. There will be people that get stuck for a while. Fibre must connect to something at its other end, all the way back to a POI. Telstra copper would only have to go to the nearest exchange and fibre transit the rest of the way. Greenfields were always going to be the hardest to do as they’re spread out all over the place, without regard necessarily to existing infrastructure.

      • Correction, if they are in an estate with more than 100 lots, NNNCo have only just sorted out this requirement. Those in estates with under 100 Telstra do.

  23. TTBoy,
    How is the NbN fiber optic network non-standard compared to the existing Telstra fiber optic network and why can’t it be interconnected with the existing Telstra network ?

    Is it the actual fiber optic cable or the transmission equipment or the actual transmission standard that NbN is using that is incompatible with the existing Telstra network ?

    Telstra is actually building part of the backbone fiber optic network for the nBn, no doubt in preparation for taking over what little is actually built of the NbN after the election.

    You certainly hit the excessive NbN bureaucracy cost dead center when you mentioned CONSULTANTS = massive ripoffs = opportunism = short term thinking only to the end of the contract = each consultant only interested in their own contract = no overall coordination or planning = chaos = exactly what is happening right now in other words a normal Labor project.

  24. Yeh. Increase my power bill by 20% but go ahead and spend all this money on some other useless junk. Yeh that’s cool.

Comments are closed.