Did NBN Co fudge its rollout numbers?

46

news Shadow Communications Minister Malcolm Turnbull this week claimed the National Broadband Network Company had in January fudged its network rollout statistics by retroactively updating its December fibre rollout database to show additional premises; a claim NBN Co has denied.

On Monday this week, after a month of requests from a number of different organisations, NBN Co released hard statistics showing the progress of the rollout and uptake of its network infrastructure during the three months to the end of 2012. The company’s quarterly numbers are being extremely closely watched by Australia’s technology sector and political elite, as they are a key measure by which the Coalition will criticise the Federal Government’s implementation of its ambitious NBN policy, ahead of the upcoming Federal Election this year.

In general, although they show only slow progress for the three months to the end of December last year, the numbers are believed to be within the guidelines of the company’s overall plan, which calls for the company to have connected some 341,000 premises to its fibre broadband network by the end of June this year, with some 54,000 live customers. However, according to Turnbull, in January NBN Co changed its definition used to show how many premises it has passed with its fibre network, in order to make its progress on its network rollout look better than it actually was.

To understand where Turnbull is coming with this claim, you need to understand the way NBN Co reports its statistics.

Most of the attention on the company’s network rollout occurs on a quarterly basis, when NBN Co releases key measures – premises passed and active services – for the three months preceding. It is this dataset which the company released this week. However, those following the company’s progress more closely have taken to keeping tabs on another document, released monthly. This database, known as the ‘Monthly ready for service rollout plan’ (available from this page), does not explicitly provide details on how many premises the company has actually connected with its fibre network, but does go into extreme detail – area by area around Australia – about how the NBN is being deployed and when NBN Co expects each region to be connected. It predicts the near future; and can be used to show the near past, by checking which areas are removed from its charts. It’s a painstaking process, but it’s there.

The database focuses around what is known as ‘Fibre Serving Area Modules’ (FSAMs). If telephone exchanges are the key central network link points of Telstra’s existing copper network, FSAMs are some of the key link points of the NBN’s fibre. If you know which FSAM your premise is going to be connected to, you can use NBN Co’s monthly ready for service rollout plan to determine precisely when the company expects to connect you to its network.

However, after NBN Co released its quarterly numbers on Monday, Turnbull pointed out that the way the company reports this monthly ready for service database had changed. In mid-December, the database was released as normal. However, in mid-January, NBN Co retroactively issued a new version of that December database containing a new category, which it described as ‘Early Access Brownfields’ – meaning areas where some residents connected to a specific FSAM were expected to shortly be connected to the NBN. You can download the original December database (which NBN Co has since removed from its site) here in Excel form, and the revised version here.

According to Turnbull, NBN Co is using this ‘partial FSAM’ model to change the way it accounts for completing construction to premises in its network.

“At least 4,000 of the brownfields premises passed have been counted as ‘early release Fibre Serving Area Modules’ (FSAMs) – meaning that the NBN is no longer waiting until entire FSAMs are ready to switch on to the network before counting them as having been ‘passed’,” Turnbull said in a media release issued on Monday, effectively alleging that NBN Co’s figure of having passed an additional 20,386 fibre premises in the December quarter may actually be closer to 16,000.

“This is despite the NBN stating in its corporate plan that the ‘premises have been passed’ once “all design, construction, commissioning and quality assurance activities in an FSAM have been completed for the Local network and Distribution network, (Corporate Plan (PDF), p.94)” Turnbull added.

Turnbull is correct in that NBN Co’s corporate plan does indeed state that the definition of ‘premises passed’ on its network entails all design, construction, commissioning and quality assurance activities in an FSAM having been completed; the corporate plan does not make allowance for ‘partial FSAM’ statistics, or, as NBN Co has defined them, ‘Early Access Brownfields’.

“This revision to the definitions (in order to make it easier for NBN to meet its targets) is in addition to having changed the way it counts houses passed in greenfields areas – it now counts ‘lots passed’ regardless of whether there are any premises constructed ready to take a service on the lot or not,” Turnbull added.

According to NBN Co, it’s done nothing wrong in adding the extra ‘partial FSAM’ data into its monthly ready for service database. Additionally, the company has not disclosed precisely how it collates its quarterly data release, so it is unclear if such information is used in the release of the quarterly statistics as Turnbull has suggested.

“The priority of NBN Co is to provide as many Australians access to superfast, reliable and affordable broadband as quickly as possible,” a spokesperson for NBN Co said in a statement yesterday responding to Turnbull’s claim. “Where we can switch on these smaller areas we will do so because we want people to enjoy the benefits of better broadband as soon as possible. Clearly we only do this in areas where there is sufficient transit and the wider distribution network has been rolled out – so premises owners can actually order a service.”

“It’s simply incorrect to state that NBN Co has changed the definition of “Premises Passed” by allowing the inclusion of parts of suburbs as well as full suburbs (known as “Partial Fibre Service Area Modules” and “Fibre Service Access Modules” in company jargon). The release of partial FSAMs in initial rollout areas was provided for in the agreement with Telstra – a document which was finalised nearly 12 months ago.” NBN Co has declined to release its agreement with Telstra in the past, and Freedom of Information requests for the document have been denied.

Upon further questioning as to why the company retroactively updated its December database, NBN Co’s spokesperson issued the following statement: “NBN Co included “Early Access Brownfields” figures in its January Ready for Service report issued last Thursday. However “Early Access Brownfields” were first included in a December update after advising RSPs at the end of November the additional information was coming.”

“We are providing this data so that [retail service providers] have an updated estimate of when services can be ordered in an area so they can plan their own marketing and sales activities.”

opinion/analysis
There’s a lot to debate about the NBN right now, and it all revolves around to what extent the company is delivering on its promises. While there are still those who are obsessed with the fibre to the node/fibre to the home debate, what we really should be debating right now as a nation and as a telecommunications industry is not what technology we should be using for the NBN, but whether the Labor Federal Government and the organisation it has created in NBN Co are actually being successful at implementing the NBN policy.

After all, despite the fact that we all know there’s a huge amount of NBN construction activity going on around Australia right now (some 750,000-odd premises under construction, if you believe NBN Co, which I broadly do), the top-line figures do not look good when taken out of their context. Since late 2007, when the then-Rudd Labor administration took power and kicked off its initial NBN policy (which subsequently failed and became the current, much broader policy in April 2009), the Government has only successfully deployed fixed broadband infrastructure to some 72,400 premises. It’s only connected some 10,400 premises for active use.

This is why I’ve focused so highly on Turnbull’s claims here. NBN Co top-line hard rollout figures are not fantastic. I think we can all agree on that much. They may be within the company’s guidelines for its plan, but they’re still not great. But at least up until now we have been able to believe that the figures, and NBN Co itself, was honest and that it was applying a consistent methodology in how it delivers those figures.

The facts are that NBN Co, since December, delayed substantially compared with the previous year in releasing its hard rollout figures, taking a month to release its key quarterly measures (the year previously it took a matter of days). During that period, it made its construction chief redundant and retroactively updated its December ready for service database to show additional premises in a manner that it had not done previously. It also kicked off a paid incentive scheme for retail ISPs to sign up new customers to the NBN.

Did NBN Co in January, take the additional step of fudging its definitions of ‘premises passed’ to show its progress in a slightly more favourable light? Right now, I don’t know. There just isn’t enough evidence to say.

Personally, I don’t believe so, because I know many of the folks at NBN Co and I have consistently seen them act ethically over the past several years. I genuinely believe the company has Australia’s best interests at heart and will attempt to act transparently and ethically in every situation, while pushing hard for its goal of ‘broadbanding Australia’. This is the culture which its chief executive Mike Quigley has instilled in it.

However, we’ve also seen some rather disturbing signs coming from the company over the past few months, including a lack of transparency and internal restructuring. I don’t think there’s a conspiracy here, but coupled with the evidence Turnbull has presented regarding its treatment of partial FSAMs, there are some questions swirling around NBN Co right now that are a little troubling.

Another motivation for writing this article is that I didn’t want Turnbull’s claims of number-fudging to go untested. It’s a serious allegation, and we need to dig deep whenever a politician alleges this kind of behaviour. Right now, I can’t say whether Turnbull’s right or wrong: But what I can say is that he was right to at least ask the question about this, and it’s to the credit of Turnbull’s team that it’s watching NBN Co’s data so closely.

What do you think? You’ve all now got the same evidence that I have: The actual documents, verbatim statements from both sides and so on. Make your opinions known. And before you comment, know that I remain in favour of the NBN policy as a whole and am not a stooge of the Liberal Party, which currently has an inferior NBN policy. However, like some others, I am beginning to question NBN Co’s performance in implementing the NBN policy. That should be the context for this discussion.

46 COMMENTS

  1. Actually Renai whilst you do raise valid points and IMO full credit to NBN Co considering the extreme unprecedented pressure they are operating under which must impact somewhat on their operations as their future existence and dedicated hard work is hanging by a thread.

    IMO the alteration actually provides a more granular and detailed metric which can only be beneficial.

    P.S my verbose comment in http://delimiter.com.au/2013/01/29/slow-progress-nbn-co-releases-dec-2012-stats/#comment-570646 re the rollout stats and remediation and bundled plans is awaiting moderation on Malcolms site, I wonder if it will ever appear?

  2. Storm in a teacup. Premises passed is the only relevant statistic for a wholesale provider anyway. Everything else is up to the RSP. You know, we are lucky to have this at all. NBNCo is not obliged to report figures to the public. To the ACCC and government sure. This level of transparency is quite refreshing. The debate over rollout numbers is getting rather boring however. It’s coming already and unless there are significant delays, which we have no evidence to date then really who cares? All you are doing is providing fodder for opposition who are largely irrelevant anyway. Tony Abbot is desperately trying to restore his reputation but its too late. Joe Hockey is just an idiot, nothing more to see here. Then there is Malcolm. Poor Malcolm Turnbull is probably missing private enterprise considering how much bullshit the coalition forces to come out of his mouth.

  3. I dont critique their rollout speed, certainly at the rate they’re going it’ll be well and truly on track. Given I dont normally go outside the NOC unless I need to, I dont have a wealth of trenching and fibre link experience.

    I would go as far as to say that they do lack staff to speed up the rollout even further, however given how much training they require it may be that there just isnt enough time, enough training facilities or even qualified staff / staff trainers to increase the rollout speed. With enough staffing or contractors, NBNCo could almost double the rollout speed.

    On Turnbull’s claim of the ‘Partial-Fibre Serving Area Module’ numbers being added – I dont see why this is a problem. Think about it. Even if the brownfields numbers are added, who’s to say they arent close to specs that NBNCo needs or similar enough that they can be directly connected now with only minimal interference or changes from either NBNCo or its contractors? I dont see why this is a critical problem, those numbers should be allowed in. If it physically isnt connected AT ALL, you cant call them within the rollout pattern; but a few minor changes and they’re connected?

    Nah – I’d call that in the lines. Much like tennis – On a line fault or an Ace; just because 3/4 of the ball is outside the line doesnt mean its not in the line. Its still ON the line, its just not perfect. :)

  4. I think I will wait for the release by NBN Co of its June 2013 figures and see what they are telling us then. My decision on who I believe will be recorded (in secret) on 14 September 2013

    Off topic but can anyone recall an election date being announced so far ahead?

    • Going to force some policy detail
      Must be some wondering if they are considering jumping out of the frying pan
      Australia will choose and DESERVE it’s future, be good if the voters have some idea of that future and how it will be achieved

  5. So if the numbers are bad then they are bad.
    If the numbers are not bad they are lying
    If the numbers are extraordinary they are lying

    Basically you could have 10 million active connections with statutory declarations for every household and they would be lying or fudging or cooking the books.

    So if you want to make accusations go to the next senate estimates hearing and actually do your bloody job and report the outcome rather than spreading unsubstantiated politicking.

    • +1 AJ

      I really don’t have a concrete opinion on passed, connected or whatever numbers “yet”, as I believe it’s too early. As such I agree with Renai’s headlines here –

      http://delimiter.com.au/2013/01/08/spectacularly-shortsighted-debating-2012-nbn-take-up-rates/

      Sure peruse them, discuss them and criticise them if you must, but keep everything in apolitical perspective.

      And as you allude to, NBNCo are, and will always be, damned if they do/don’t in every facet anyway, in the eyes of the detractors… and their contradictory comments reflect this pretty well daily.

      The most glowing contradictory hypocrisy in relation to this topic, being IMO, them scoffing and rejecting the Corporate Plan projections as nothing more than unfounded estimations (well der – WTF do you think all construction estimations everywhere are) and as such claiming the Corp Plan is of absolutely no significance and should be dismissed…whilst simultaneously expecting NBNCo to reach or outdo NBNCo’s own target estimations :/

  6. Thank you Renai LeMay.

    It is so refreshing to see an issue like the NBN reported by a well informed source and presenting a balanced point of view. I haven’t taken much notice of the NBN chatter in the past, but after reading this I really do feel more informed and aware of the issues, and able to form my own opinion on the issue.

    Great journalism !!

  7. Havent some quarters of the media been calling for monthly reporting to get a more precise idea of where things are at? This, for all intents and purposes provides that level of detail. And yes, that is (in part at least) refering to you Renai.

    If it takes 12 months to roll out an exchange, for that 12 months the properties are going to stay as a stat in the “under construction” stat, then become a big bump in the “completed” stat. This smoothes that bump out, adding the good to go stats as they are good to go.

    If the properties passed arent capable of actually connecting, then its a bit of a furphy, and one I’d be disappointed in, but in some ways this is just giving you what you asked for Renai.

    Look at it from that perspective. While it isnt speeding up the reporting of stats, it IS fine tuning the detail.

    • “This, for all intents and purposes provides that level of detail.”

      No, it does not. For starters, there’s no active services listed …

  8. No, I don’t think this extra information isn’t fudging the figures, because as long as the premises passed can get their NBN, that’s what the rollout figures are all about. From what I can see, that’s the case, so the numbers stand and can be checked anyway if someone is suspicious.
    Previous media releases have included the partial FSAM figures so these figures must have been able to be sourced previously, they are just in the public domain now.
    The only reason these figures are probably relavent is that the NBN has not only to roll out it’s network but be seen to be rolling out it’s network. Hence the partial rollout figures, along with all the other detailed reporting.

    So summay, the extra information isn’t fudging the figures.

  9. “Did NBN Co in January, take the additional step of fudging its definitions of ‘premises passed’ to show its progress in a slightly more favourable light? Right now, I don’t know. There just isn’t enough evidence to say.

    And yet, here’s Turnbull falling back to the old favourite line so often used by Abbott — “cooking the books”.

    It would matter little if the numbers were poor, or stellar; Turnbull is never, ever going to have a positive view on the NBN. He is not permitted to.

    Apart from some assumptive reasoning from the Wizard of Wentworth, is there anything to actually back up the claims?

    • There’s a great big question mark!! (please forget that Malcolm made the question mark and sticky taped it onto the NBN building).

      When they answer the question; the question mark will be unstuck; the answer ignored; and another question asked and the giant question mark will be stuck on NBNCo again, the great big government book cooking liar that never tells anyone any answers to any questions ever!

  10. “delayed substantially compared with the previous year in releasing its hard rollout figures”

    The previous year there was a fraction of the data to collate surely.

  11. Also, the answer to the topic is:
    No.

    If you want more detail, the answer is:
    What was the question again?

    One reporting metric != another reporting metric.

    monthy ready for service IS NOT the quarterly report.
    Adding additional info to the monthly ready for service; (IE telling RSP’s that half the customers on one FSAM are ready to go) does not mean that *all* your figures have been changed.

    Oh it might, it absolutely might. But in this instance the entire story should be:

    NBN Figures, I am seeking clarification.
    I have noticed (or the big MT’s intern has noticed) that the figures in the monthly ready for service document include some new information. I have asked NBN co if the same “half complete” FSAM data is included in their “premises passed” figure. I have also asked NBNCO to clarify if it has *always* included such figures. And I have asked NBNCo to clarify the statement in the Corporate Plan where they explain what constitutes “premises passed”.

    Notice how the paragraph above jumps to the minimum of conclusions? Notice how it clearly indicates what the controversy is; all without “cooking books” “fudging figures” and other hyperbole. *sigh*.

    • But where is the outrage to get people coming back and posting.

      Unfortunately the truth does not stimulate debate this article will get over 100 posts in the comments your article would get 20.

    • hey PeterA,

      for the record, this is precisely what I did ask NBN Co. The two responses I got back are in the article above verbatim.

      Cheers,

      Renai

      • Are the partially completed FSAM’s able to be connected to customers? Do they completed connections have the same functionality as those on a fully completed FSAM?

        To me this would impact strongly on whether the statistics can be included or not.

  12. I’m obviously confused here, because it seems to me that Turnbull is saying they’re cooking the books* because they’re including numbers from partially completed areas. What is wrong with that?

    *”cooking the books” refers to finances only, surely?

    • Are finances the only area in business that require you to keep detailed numerical records?

      I see it as fudging (doctoring) the accounting process for recording keeping for anything in the business environment e.g. inventory, hours worked, finances, anything that you need to keep record of.

  13. I dont think Turnbull can be trusted to properly represent the figures presented to him. He has regularly misrepresented other data and facts provided to him about NBN Co funding, Quigley not taking bribes, etc.

    Why should Turnbull suddenly be trusted to be truthful and forthright about this?

  14. I don’t really care how they arrive at the figures as long as they publish targets in advance and then report actual numbers on the same basis as the targets. That will allow us to track their progress against their plan.

    • I agree with this.

      Recently I have seen some NBNco targets for the next two quarters. I will be keen to see how they play out.

  15. so let me get this correct, nbnco hand out much more detailed data, which people have been calling for, and because there is now data from partially completed areas (ie more up to date/accurate data) its ok for turnbull to accuse nbnco of cooking the books?

    do you want more accurate data from them or not? how about we tell them no thanks, we want the same stuff youve been handing out before, you happy with that?

    and i really like the bit where while you dont believe that (fudging the numbers) to be the case youll add a *but* in there anyway to keep this sort of crap alive? stop attempting to sit on the fence so damned hard, its pathetic. there is nothing wrong with reporting accurately no matter what the outcome is, you dont have to force in views from the opposite side just to balance it out for appearances sake.

    btw, nbnco do not change their reporting metrics all by themselves, they report specifically what they are told to, so please get your terminology right – wouldnt want you spreading lies that nbnco is a rogue operation doing whatever it wants regardless of instructions from its sole shareholder now would we.

  16. WTF??

    How is including partially completed areas that are able to ring up and get the NBN connected “cooking the books”? They are premises passed that can order the service…FULL STOP!

    In the area my sister lives there are streets from her FSAM that have already gone live despite the expected completion of the FSAM not being until March. You can actually see the little blue pockets where this happens on the map.

    But just because it improves the “premises passed” numbers for NBN all of a sudden Malcom starts mouthing that they’re cooking the books just because he needs something to bitch about…

  17. I was originally not a supporter of the use of the collective connections commenced or completed statistic (though not quite as opposed to it as Renai) however yesterday’s results completely changed my opinion because that figure is in addition to other figures whereas I had previously believed it to be the main figure provided (i.e. no premises/lots passed figure would be provided). It gives us an idea of whats going on and how much work the contractors are doing and the more figures we have, the better.

    Whilst the numbers were nothing to be ecstatic about, your rant was actually quite a surprise to say the least. These numbers demonstrated that NBNco was indeed ramping up progress.

    Yesterday’s headline “slow progress” was something I’d expect from The Australian or AFR. What we saw was steady progress, not slow progress – steady progress but there’s still a long way to go.

    Also, the fact that the election will be held on 14th September reveals that the whole point of the $108 incentive is not due to low adoption rates but to increase adoption rates and figures right before the election. At least that’s the interpretation thatmakes most sense to me. i.e. Its entirely political and not a failed business model.

  18. If there are premises at which an NBN connection could and would be made as soon as the owner of the premises so requested then its totally reasonable to count it as a “premises passed”.

    Its irrelevant that some other premises in the same area could not get a connection because the cabling to them isn’t operational yet.

    But it is relevant if there are not and never have been any premises there. Empty blocks are not “premises”. If there is somewhere the premises end of the NBN connection could be placed then its a “premises”.

  19. Good on you Renai for putting these type of questions out there. No doubt when answers are forthcoming you can either expose the NBNco for being a bunch of liars, or confirm for the umpteenth time that Turnbull is one.

    If unchallenged such questions can hang over the head of NBNco, all the way to the election. The fact that we are left with a simple allegation from one side and a denial from the other, without any evidence to know who is right or wrong, is obviously not a great place to be.

    • You don’t require to know the numbers, since the project has a time frame.

      These numbers are used for political nonsense for short term political game and long term tax payer pain.

      Recently Malcolm said most if not all brownfield development of FTTP will be gone.

      And the only ones getting FTTP will be greenfield (his mates).

  20. Once Again I voice my concern to the Media and Malcolm.

    For one, Malcolm who likes to really fudge the numbers, since he’s the Opposition Member and recently gave the NBN a kick up the arse after the election.

    Two, the media for actually might be maybe believing the numbers…

    I for one don’t give a shit.

    Roll on NBN.

  21. Whilst they may have added a new category, I think the accusation of fudging the book is usual Coalition hyperbole. It appears, to me at least, that the addition is actually resulting in a more accurate measure of homes passed which is a good thing surely?
    I guess I can understand why the change would be disgruntling for the ‘criticise the NBN at all costs’ group of people, as it definitely skews the number in NBN’s favour by reducing the time lag in the reporting of houses passed in the early stages of construction. However, a reasonable person can surely see that as long as the requirement that a house is actually passed by activated fibre to be counted, then the new method is of higher quality than the old method.
    The logical solution is for the published targets to be updated to the new counting methods, which I would imagine NBN will shortly be doing, if they havent done so already.

  22. So if NBN Co are release partial FSAMs in the MRFS report, that seems to imply that partial FSAMs will be ready for service before the full FSAM is. Does this partial activation apply only to the

    Allowing people to have active NBN connections (Renai’s metric of choice) as soon as their FDA completes instead of waiting for the whole FSAM to be connected seems a good idea to me.

    Aside: Jan 2013 is not the first time NBN Co has revised MRFS figures (eg Nov 2012). On top is this, the 1y plan (rollout-info-1-year-july-2012.pdf) also has a v2. As for the delay in releasing stats, the Jan 2012 report had considerably less data to collate than the Jan 2013 one (that said, I do agree that it should be released as soon as possible).

  23. the NBN is going to start counting homes that have fiber to home connections in places like new estates next even though the NBN had nothing at all to do with those homes getting a FTTN connection

    its like a big scam

Comments are closed.