news The National Broadband Network Company has labelled claims by iTNews yesterday that it was deliberately manipulating its rollout figures in order to put a more positive light on them as “entirely untrue”, and argued that the publication’s claims are “not supported by the facts”.
Yesterday the publication published the results of what it said was a two-week investigation into NBN Co’s rollout figures. It stated: “NBN Co has continually promoted the impression of progress by announcing that its fibre rollout has expanded to new towns and cities. Today iTnews can reveal that NBN Co has resorted to connecting fewer premises within existing sites than originally promised to allow for this expansion.” The Australian Financial Review subsequently published a follow-up article stating that it had “verified” iTNews’ analysis. Delimiter has announced that it will over the next week publish a fact-checking analysis of the claims.
In a statement issued last night, NBN Co’s executive general manager of planning and design, Peter Ferris, said iTNews’ assertion that NBN Co was deliberately manipulating its figures was “entirely untrue”.
“A correct interpretation of the data published on NBN Co’s website confirms the following facts: No previously announced construction modules have been removed from the construction program; No areas are “losing previously-promised fibre connections”,” said Ferris. “Furthermore, ITNew’s claim that “towns and cities have disappeared” from the rollout is not supported by the facts. It looks to be based on an erroneous assumption that FSA boundaries are fixed over time.”
Ferris said that iTNews had given the specific example of the Townsville suburb of Gulliver being “one of the biggest losers from the redistribution”. “Yet Townville has not lost out at all,” he said. “NBN Co received details that the size of the telephone exchange at Gulliver wasn’t large enough to serve 12 [fibre serving area modules or FSAMs]. So we had to transfer 14,000 premises from the Gulliver FSA to the Townsville FSA bringing its total FSAM count to 12 and reducing Gulliver to 8 FSAMs.”
“To reiterate: In the original planning, Townville had 8 FSAMs and Gulliver had 12 containing approximately 50,000 premises. In the current plan, Townsville has 12 FSAMs and Gulliver has 8 – still containing 50,000 premises. This is merely one of many such errors in the report.”
“In fact,” added Ferris, “between the January and February Ready for Service releases there was also an update to our One Year Construction Rollout Plan. This plan included the movements of a number of FSAMS from one [fibre serving area] to another as we applied the network design rules. These rules take into account factors like the maximum length a fibre cable can run, or the availability of existing infrastructure, and they are applied when NBN Co moves from a “desktop (PC based) design” to a “physical design” – based on on-the-ground site inspection.”
“The effect of the changes were published in the February Ready for Service data and resulted in changes to the number of FSAMs in some FSAs, which affected the premises count in some areas.”
Furthermore, said Ferris, there had also been some fluctuations in the numbers of premises in different construction modules reflecting the fact that more detailed information about premises counts was uncovered as NBN Co moved from desktop design to detailed physical design following area-by-area walk-outs.
“However, these changes in premises in particular FSAMs do not alter the construction target,” he said. “NBN Co is still working to the target of passing 286,000 premises by the end of June this year – and the spreadsheets referred in the article actually indicate this.”
“NBN Co’s construction progress is monitored and managed on a daily basis and the company issues a monthly “Ready for Service” spreadsheet report to service providers with forecast dates of connection. Dates for FSAMs move backwards and forwards in this report depending on real construction progress, which can be affected by availability of connecting infrastructure, severe weather events such as floods or bushfires, or the readiness of underground conduits.”
“However, the forecast date for FSAMs to be ready for service can – and do – get pulled forward after having previously moved backward. What the figures show is how open and transparent NBN Co is about the rollout, even down to the expected ready for service date for individual FSAM or some even smaller Fibre Distribution Areas.”
With respect to the issue of new developments (greenfields areas), NBN Co head of business operations Kat Stapleton added the following statement:
“NBN Co has previously said that the Greenfield build is one of the most challenging aspects of the rollout, as it is based on demand from developers for works that are spread right across Australia. Connecting new developments is a particularly challenge in areas where NBN Co hasn’t got existing network to connect premises to, and the rollout relies on a number of critical factors such as access to pit and pipe infrastructure in which to install. We are extremely focussed on the need to provide services in areas where people have already moved in, and we are applying additional resources and effort to provide these services as soon as we can.”
iTNews has published a new story containing NBN Co’s response but stating that “our story holds true”. “NBN Co’s statement ignores the central theme of our analysis: that parts of the rollout are slipping or delayed,” wrote the article author Ry Crozier. “NBN Co’s statement does not address these delays. It evasively argues that the story is incorrect because we said these connections had ‘disappeared’. It’s a play at semantics that doesn’t hold any weight against our analysis.”
In its response to NBN Co’s statement, iTNews accuses the company of ignoring the central theme of its analysis, that parts of the NBN rollout are slipping and delayed. However to my mind, NBN Co has precisely addressed these issues.
It is obvious at this point that there are parts of NBN Co’s rollout that have been delayed compared to where they were originally planned to be at this stage. There are sites in Western Sydney where this has occurred, due to what appears to be difficulties between NBN Co, Telstra and local councils with respect to duct remediation. There are whole states suffering issues at the moment due to some issues with NBN contractors such as Syntheo. All of this is on the public record.
However, that wasn’t what iTNews alleged. It alleged that NBN Co had been deliberately manipulating its rollout figures in order to put a more positive light on the rollout.
I haven’t looked in detail yet (as part of our fact-checking exercise) at iTNews’ analysis in terms of the raw data. However, the statements made by Ferris last night with regard to these issues being a normal part of the cut and thrust of such a large national fibre rollout ring very true. Close observers of NBN Co’s work have known for a long time now that there is a significant difference between the way the rollout looks like on paper and what engineers and construction crews encounter when they’re actually out there in the field doing the work. NBN Co has given a series of extremely plausible replies to iTNews’ analysis — including examining specific examples.
In addition, consider the sorts of questions which iTNews sent NBN Co last week (at the end of the week). The publication lists one example on its site: “Why did the approx number premises for each FSA change across the board in February – often the shift per FSA is a matter of a couple of hundred premises between old and new e.g. Blacktown’s 28,300 becomes 28,600, but with three less premises per FDA for 2BLK coded builds (just as an example). Did NBN Co change its methodology (otherwise what reason would there be for such large scale shifting in those numbers)?”
These are detailed questions, but they don’t appear to go anywhere near directly accusing NBN Co of deliberately manipulating its rollout figures or alerting the company of the gravity of the article which NBN Co was considering at the time. Perhaps if the publication had taken a different approach, NBN Co would have been more timely in its response and filled in some of the blanks before iTNews published its article.
The work iTNews has done is admirable. It’s great that this kind of deep analysis is being conducted on NBN Co’s data. However, it appears as if the publication failed in two critical steps. Firstly, although it conducted two weeks of research on NBN Co’s data, it failed to speak in depth to the company to source context and insight from real human people into the cold data which it was analysing. This context — which Ferris went a long way to supplying last night — might have totally changed the character of the original iTNews story.
Secondly — and this is the part I personally consider unforgivable — the publication did not provide any direct evidence for its claim that NBN was deliberately fudging its numbers. This claim cannot be backed up by statistical analysis alone. It requires that iTNews had human sources that could verify human intent to deceive.
This is not a trivial issue. It’s why professional journalists are often willing to state that a politician, for example, appears to be misleading their audience, but are very seldom willing to state that a public figure is flat out “lying”. Making that statement requires a higher burden of information — direct knowledge of that person’s thoughts and motivations. In this case, iTNews appears to have relied almost solely on its statistical analysis and has discounted the human factor. I cannot approve of this approach.
A number of commentators on iTNews’ own articles on this subject have in fact made this precise point.
There is also a broader question as to what extent the analysis is actually relevant. NBN Co’s pledge to the electorate is not to have deployed fibre in a certain area by a certain date. It is to have finalised a certain number of premises by certain dates (for example, 286,000 by July 2013). NBN Co is free to prioritise its rollout within those boundaries as much as it likes; in fact, operations in the field will require this flexibility as certain conditions come up in certain areas, as Ferris discussed. There is currently no indication that NBN Co will not meet its July 2013 targets. In fact, indications are that the company is staunchly on track. I encourage those commenting on the situation to remember this fact.
In my view, with its response, NBN Co has left iTNews looking a little naive and even childish in this encounter. In my opinion, iTNews should have spoken to NBN Co in a great deal more detail about this issue before it went to print, and it should have realised that you cannot base a claim of deliberate manipulation on statistical analysis alone. But then, that’s not really the way Australian technology journalism is going these days.