Note: Where available, the PDF/Word icon below is provided to view the complete and fully formatted document
National Broadband Network Select Committee
04/03/2016
National Broadband Network

HOGAN, Mr John, Head of Technology, Stan Entertainment

LORKEN, Mr Andrew, Executive Director—Broadband, PMO, IT, Foxtel Management Pty Ltd

MEAGHER, Mr Bruce, Group Director, Corporate Affairs, Foxtel Management Pty Ltd

SNEESBY, Mr Michael, Chief Executive Officer, Stan Entertainment

[14:15]

CHAIR: I welcome representatives from Stan Entertainment and representatives from Presto and Foxtel. For my edification, Presto is?

Mr Meagher : Presto is a joint venture between Foxtel and Seven West Media. We represent—

CHAIR: Presto and Foxtel?

Mr Meagher : Yes. To the extent that anything is relevant to Presto we are happy to talk about it, but we would also like to talk more broadly about IP-delivered services.

CHAIR: Do both groups wish to make a brief opening statement before we go to questions?

Mr Sneesby : I thought I would provide a bit of context for the purpose of the discussion today. As background, Stan is an SVOD streaming service. We operate in the same category as the likes of Netflix, with the key differentiation for us being our content line-up—a differentiated content line-up to others in the category. Recently we completed the first year of our operation, so we have just passed 12 months since we launched the service. During that period of time over 750,000 households have signed up to the service in Australia, and we currently have an active subscriber base of over 400,000 Australian consumers. Customers pay Stan a $10 monthly fee for full access to our catalogue, which is over 8,000 hours of television and movies, and our service is delivered 100 per cent over open internet.

In terms of today and where we are as a category, we estimate that the Australian SVOD market size is just over a million households, with—

Senator LUDLAM: Can we just hit you up on that acronym, for people who are playing at home?

Mr Sneesby : Sorry, yes. SVOD is subscription video on demand—that is, accessing, as I explained before, that complete library of content for a single price per month. And it is delivered over the internet using streaming technology, so people quite often refer to streaming broadly when they are talking about the category.

As I said, we estimate that the market size in Australia today is just over a million Australian households, with many of those Australian households taking up more than one subscription video-on-demand service. Our product today supports streaming in high definition across all of our platforms, and to use Stan the average household needs as an absolute minimum of two megabits per second of bandwidth to get reasonable access. We assume multiple users of the service in a single household would need approximately 10 megabits per second of bandwidth and then, ideally, for full HD across multiple users in a household, around 20 megabits per second. Today's broadband infrastructure, as we see it, in a lot of cases is adequate for the majority of streaming households, but this is largely due to the use of what we refer to as adaptive bit rate technology. What that technology does is adjust the quality of a streaming service downward to make up for low bandwidth availability. So if you look at the average bit rate that our audience sees, you will see it is around four megabits per second. What that equates to for the consumer is around the low end, like 720p high definition. The low end of the high-definition stream is what the average consumer across our network is experiencing today.

There are two key points which I think are relevant for our business. Firstly, there is obviously a percentage of the Australian population that does not have access to that minimum amount of bandwidth of two megabits per second. Obviously, that means that they are unable to access a service like ours. Then there is a larger proportion of the population that does not get access to the full HD experience that you might want to experience with a streaming service. Secondly, we observe a level of last mile congestion during peak times—peak times for us being in the evening, between 8 pm and 10 pm. What we observe in the peak period when compared to the hours just prior to the peak period is around a 50 per cent increase in the proportion of users who are experiencing a degraded service. From the early evening hours to the later evening hours there is a 50 per cent increase in that incidence of degradation, and around a 30 per cent increase in those consumers who are experiencing what we call buffering, which is effectively pausing in their video during that peak time.

Looking to the future, we estimate that the market in Australia will grow to four to five million households within about four years. So by around 2020 we expect the take-up to be around four to five million households, or around 50 per cent of the Australian household population. During that period of time—in that time frame of three to five years—we expect that the mainstream will begin to demand ultra HD, which is referred to commonly as 4K. It is something that is available today, but it is somewhat limited by the availability and expense of ultra HD or 4K television sets, as well as the availability of content which is produced in 4K. But over that three- to five-year time frame we expect that consumer demand to be quite mainstream as those television sets get cheaper. Virtually every TV within the next 12 to 24 months that is sold at retail will support 4K. More and more content is being produced in 4K. Today you require around 25 megabits per second to support a single 4K stream. The technologies that we are currently looking at are expected, within that time frame of three to five years, to drop the requirements for 4K to around 12 megabits per second. So we are expecting to see compression technologies on our side meaning we can deliver a 4K stream to the household for around 12 megabits per second rather than 25 megabits per second.

What that does mean for our future needs for bandwidth is that at the bottom of the rung there is still a minimum entry point of two megabits per second for consumers who are receiving a low-resolution stream just to be able to get a stream to the household. For those who want to consume 4K streams with multiple devices in a household, you will be looking at a range of between 20 megabits per second and 40 megabits per second for a reasonable multi-user household. At Stan we currently have rights for a range of 4K content and we are shooting our own original productions in 4K, so that catalogue of content is building. We would expect to launch a 4K service some time in the future. I hope that gives you some context.

Mr Meagher : Foxtel is very happy to appear today to assist the committee in its current investigation into the internet capabilities needed for on-demand streaming services. Foxtel's services are increasingly provided via IP as consumers change their viewing habits and expectations. In short, they wish to watch what they want, when they want and on the device they want.

We expect the underlying trends driving the need for faster access speeds and greater network capabilities to continue. Faster access speeds will be required to support high-definition video, often streaming concurrently to multiple devices in the home. Network capacity will need to grow as more homes adopt and consume IP video services. In particular, mass-viewing events that encourage concurrent streaming across the community, such as major live sports, will have a significant impact on capacity when delivered over IP. Networks' performance will also be affected by significant spikes at peak hour downloading times, such as weekend afternoons—which, again, is largely driven by sport—weekend evenings and other family entertainment times. Data used in peak download times is increasing far more rapidly than at other times, and this has implications for pricing models.

Foxtel's 2.9 million customers receive Foxtel content via IP each time they use Foxtel Go, which delivers linear and on-demand services to mobile devices; log on to our made-for-streaming service, Foxtel Play; download a movie or a TV show on demand via Anytime on our set-top box; use enhanced set-top box features, like Start Over and Look Back; or stream content on Presto which, as I mentioned, is a subscription video-on-demand service which we operate as a joint venture with Seven West Media.

Our most advanced set-top box, the iQ3, moves seamlessly between broadcast and IP modes, depending on the service being delivered. Similarly, we use adaptive bit-rate streaming and progressive download delivery methods to enable content to be delivered over internet connections that vary in quality and speed.

In the last 12 months our total volume of IP-delivered video content across all services has grown by 86 per cent. Data usage associated with Presto grew by 121 per cent over the December and January period alone. This was largely driven by the launch and heavy promotion of one single program, the Home and Away movie. This illustrates that data usage can be highly variable and is difficult to predict. On average, we expect total data delivered to our customers via IP to double each year for at least two years and possibly to accelerate after that.

To service this we anticipate that our content delivery network costs will triple in the next two financial years. Along with this, we will be investing—as Mr Sneesby has also indicated—in improving IP delivery and compression and encoding technologies, with the aim of facilitating continued improvement in products and service and achieving overall efficiencies for the delivery of larger and larger quantities of content.

As I have mentioned, and as Mr Sneesby did as well, video consumption must be seen increasingly as a per device and per user concept, rather than simply looking at a home. We estimate that in the average Foxtel home, streamed video content is being viewed on at least five devices, often at the same time. As more of these devices support HD video and as the video content they access is delivered increasingly in higher-quantity formats, it is inevitable that bandwidth requirements will continue to grow rapidly.

The rollout of NBN is expected to significantly reduce the technical barriers to quality video delivery, with the majority of households able to support the required number of HD streams by 2020. However, from a commercial and innovation perspective it is critical that the NBN pricing model is flexible enough to enable ISPs to economically provision their services to meet consumer needs while maintaining quality. Current NBN pricing—particularly the CVC component—is applied in a relatively rigid fashion. Charging is tied to peak usage, rather than to average or some other blend. As I said, peak hour traffic has been growing at a greater rate than average traffic and is forecast to continue doing so. Therefore, the cost to ISPs of all bandwidth usage will continue to rise irrespective of varying usage patterns for different customers and at different times of the day.

We are pleased to note the recent comments of nbn co at an estimates hearing. They acknowledged that traffic growth has increased faster than expected and that they are going to review their CVC pricing construct. It is important that as the network continues to be rolled out its pricing model is set to support increasing traffic at manageable cost to the ISPs so that they in turn can maintain their quality of service and deliver not just video content but other important online services as well.

This will also ensure that end users are not exposed to significantly higher bills and that there are no artificial constraints on the innovation facilitated by the nation's move to high-speed internet.

CHAIR: Can I just interrupt you there?

Mr Meagher : Yes.

CHAIR: What is a CVC? I am new to this area!

Mr Meagher : I will refer to my colleague.

Mr Lorken : It is virtual circuit pricing. So nbn co charges an access fee, which is an AVC, and then the bandwidth or usage used is the CVC charge.

CHAIR: So it is volume?

Mr Lorken : Yes.

Mr Meagher : And it is the critical component in delivering these sorts of services.

CHAIR: Thank you.

Mr Meagher : I only have one last comment, which is that in undertaking their pricing review we would encourage nbn co to have an eye to the different requirements of ISPs which, like Foxtel, take their service from a wholesale supplier rather than directly from the NBN itself. Thank you.

CHAIR: Thank you—sorry to interrupt you. I did not realise you were close to the end there.

Mr Meagher : That is okay. There is a lot of jargon in this!

CHAIR: Tell me about it!

Senator LUDLAM: It is great that you fellows are here this afternoon. I guess that you are a bit on the front line, in that capacity constraints on the network impact on your business models directly. If you do not want to get into the politics of it, I would be more than happy. But can you lend us your advice on what you think the priorities for broadband policy should be for this parliament coming into the next electoral cycle? Maybe to frame this question a little bit, all four of you are in the private sector delivering a retail service to customers. What do you think the public sector should be doing—nbn co, in particular?

Mr Sneesby : Are you talking specifically in relation to speeds, access and availability?

Senator LUDLAM: Sorry; hardware.

Mr Sneesby : We do not operate so much on the wholesale-retail side of things because we operate over an open-internet model. But, as I noted today, I think there are two broad challenges. They are a challenge, obviously, for our commercial business model but that also means that they are a challenge for consumers getting access to the kinds of services that they want at a quality level. If I look at the volume, we see on an ongoing basis—not a high volume necessarily given that it is a small percentage of the population—certainly a consistent level of consumers who are upset that they are unable to access our service in their particular area because of the lack of availability of bandwidth. That issue and addressing that in the sub-2-meg regions is certainly something that is important for our consumers—the consumers of our product. Then, more broadly, as I mentioned before, the consumers who are able to access the service but are not able to access it at the highest level of definition, which is high definition for us, is addressing consistency of the kind of bandwidth requirements that I mentioned before. That is in the range of 10 megabits per second up to 20 megabits per second in larger households with more devices. I think in terms of what we see coming back from our own customer base and from our own audience, today those are the most immediate aspects that we see on a regular basis.

Mr Lorken : We would echo that as well. I think that is probably priority No. 1. I do not think the nbn, or anyone else for that matter, understood the pre-Netflix, pre-Stan, pre-Foxtel IP world versus the world we are in now. It strikes me that probably the most important assumption that nbn needs to make is bandwidth growth over the next few years to 2020. If they get that assumption in their business planning cycle incorrect, we do not want to see a situation where by 2020 the entertainment industry is 'competing with', as Professor Gray mentioned, e-health and education and other high-bandwidth consuming industries—certainly, nobody wants that. The importance of assumption is probably the most important thing for us, and hopefully for nbn as well. From a policy point of view, to answer your question directly, we would love to see that assumption made public and, in fact, be a collaborative process to get that number correct. Bruce and Stan have mentioned the quite massive growth rates and, whilst they may taper off in the near future, we still expect quite large data-growth rates, and then by 2020 with e-health and education coming in as well, we would hate to have a situation where nbn and the RSPs are having to monitor and somewhat manipulate the network in order to deliver high-bandwidth services and then be forced into a situation to prioritise those services.

Senator LUDLAM: You would be some of the best people in the country to tell us what you think those growth expectations should be. If I can push you out beyond a four-year horizon because this is a piece of infrastructure that is going to have a lifespan a lot longer than four years. Maybe take this as two separate parts if you like: what are your traffic growth projections out to 2020, and can you see anything on the horizon that would indicate that the growth rate is going to slow after that?

Mr Lorken : From a pure number of terabytes, in other words the amount of data that will be flowing through as traffic through the various RSPs to Foxtel customers, we see a six-fold increase in the number of terabytes between now and, in fact, the next two years. We see that tapering off possibly through to 2020, but you are still in three- to four-fold increases just in the sheer amount of data that will be pumped through each and every year to our end customers. As Bruce mentioned, we have seen an 86 per cent growth rate. We see that being consistent for the next year and then possibly tapering off, but we are certainly estimating 40 per cent to 50 per cent data growth rates through to FY19 and FY20.

Senator LUDLAM: So your most pessimistic estimate there was a doubling every 24 months?

Mr Lorken : Pretty much. That is right. Obviously, we are keen to compare that to nbn's assumptions to ensure that—from an aggregate point of view across all industries—their assumption is somewhere in the vicinity of that.

Mr Meagher : The other input—which we have mentioned in passing, which you also will have to take into account—is what happens at our end in terms of compression and other technologies that can improve. It is the interaction of those two things—available bandwidth and—

Senator LUDLAM: Cleverness.

Mr Meagher : Cleverness. Yes, exactly.

Mr Sneesby : I think it is important to look at the two aspects of what we are talking about here. One is the pure volume of data that is being pushed over the open internet, and that is a function of all of our subscriber bases growing. If we look at that pure volume aspect of things, we would also expect to see at least doubling of the volume that we are going to push over the internet or that subscribers are going to suck down from our servers over the internet each year over the coming three, four, five years. The other component is that last-mile speed. We are talking about pure data volume more than doubling every year, but then we are looking at the speed requirements to the household at the end of the line. That was the other aspect that I spoke about.

The requirements for speed will not increase at the same rate as the data volume is increasing. The categories are growing extremely quickly and driving a lot more data, yet it is the technologies at the end of the line—that shift from HD primarily to 4K over the next 3 to 5 years—that will approximately double the bandwidth requirements in that space. Beyond that period of time—beyond 2020—there are already discussions and standards being developed for 8K. The next evolution from 4K to 8K which, as you might expect, is a greater level of definition in video streaming that requires, again, a greater level of data streamed in order to support it. It is difficult to say whether that becomes a reality and over what period of time, and whether it becomes a mainstream technology. It is kind of beyond the planning cycle that we look at today.

Senator LUDLAM: Between the four of you—your most conservative growth estimates and your most optimistic estimates of compression cleverness—when do we smash into the ceiling that would be locked in with a fibre-to-the-node build where you are, whether you like it or not, locked into copper for your last mile?

Mr Lorken : Certainly, from a Foxtel point of view, we do not have the access technology experts to really know where that 'ceiling' part of the equation is. We certainly have our demand growth assumptions, but we were hoping it would be more collaborative with nbn in terms of working that formula out—or colloquially, seeing where the two lines of the graph meet.

Senator LUDLAM: There were reports in the press in the last two weeks that the lines on the graph meet somewhere between 2020 and 2025, which sounds reasonably consistent with what you are telling us today. You folks are in charge of one of the sectors where it is probably the easiest to estimate at least the next couple of years of demand growth, so you are actually quite a helpful metric. Does that sound about right, or would you not want to backup those kind of estimates? Mid-2020s. You simply start hitting the capacity bottleneck at the last mile for streaming services.

Mr Sneesby : As I say, I am not too familiar with what the capacity projections are. We did see the publicly stated objective of, I think, reaching 100 per cent of households with up to 20 or 25 megabits per second by 2020 and then beyond that, 50 megabits per second at around 90 per cent of households. I think, certainly, in that time frame, it is consistent with the needs of our business. Beyond that, as I said, it is difficult to say. I think the only thing that is really likely to drive greater bandwidth is if the shift happens towards 8K and that becomes a really viable commercial need within the household, then that is certainly going to have an impact on bandwidth requirements.

CHAIR: Mr Lorken, you said a couple of times today you would like to be more collaborative. Can you give me an understanding of the way nbn talks to yourselves—to all of the potential users—so that there can be a shared view about what the growth in usage or growth in opportunity really is? How do they talk to you?

Mr Lorken : Just to clarify, I certainly do not mean it as a criticism of nbn at all, so thanks for the question to enable the clarification. We have an account manager from nbn that we have held regular discussions with, and that is very good. It is where they draw the line in terms of commercial in confidence and being unable to share things like their assumed data growth. Of course they may or may not be able to share that, and that is okay, but I would like to see more transparency in the process so that they are collaborating not with just us and Stan and Netflix and others but of course with e-health and the other industries that will be high bandwidth users and then forecasting as accurately as possible. The importance of that, again, is so essential—it will go into the way they design an architecture for that network and what access technologies they deem appropriate, and of course it will go to the commercial model that is used as well, which is obviously a direct flow-through to the RSPs and then ultimately our consumers.

CHAIR: So they are asking you what your projected growth might be. Do they test that?

Mr Lorken : I do not know, it is a good question. They certainly should be able to test it to see what bits and bytes go through the network, but of course there is the RSP layer before you get to the ultimate subscriber so it really depends on their operating systems to see how they can delve into their network. I am unable to know.

Senator LUDLAM: I guess that is the thing that interests us most and where your business model collides with the realities of the network that is actually being built out. If you are just talking about the 720p service, I guess the minimum standard of quality if you are trying to sit at home and watch a movie, like you were telling us about, is 4 meg per second. In terms of potential market size for a 720 PBO stream at peak hour, at eight o'clock at night, in the city, or wherever you are trying to watch it, how many people in 2016 are excluded from being able to access, or not even bother taking up a subscription, of the kind of service that your companies offer?

Mr Sneesby : One of the things it is difficult for us to assess is who is missing out. We obviously can look out into our network and see what has been delivered to the consumers who subscribe to our service, so we can get an understanding of that. We can pull data—I do not have it at hand today—from our network to understand who from our customer base receives a lower than 720p service—I am happy to provide that data separately if you would like it—but what we cannot see, obviously, is outside of our consumer base how many people might be restricted from that or how many people may not sign up to a service because they are aware that their internet is not suitable for the service. We can give some information on what it looks like within our customer base if it is helpful.

Senator LUDLAM: I think it would be helpful because obviously what we are trying to do is roll this whole conversation forward five or 10 years, say to 2020 or to the end of the NBN build in 2021, or whenever they actually land it. At the end of that build cycle how many people will still be excluded say from a single 4k stream into a home, or what happens when you have five devices try to stream at the same time? Without breaching commercial confidentiality, obviously, anything you can tell us about how you view the size of your market now and in four years time, say—this must be your bread and butter, really—according to different expectations of what nbn co is going to be able to deliver, would be helpful. Do you each have customer service departments that handle complaints?

Mr Meagher : Yes.

Mr Sneesby : Yes.

Senator LUDLAM: Presumably people might sign up to your service without realising that the broadband service cannot handle delivery. How frequent is that?

Mr Lorken : I would have to take that on notice to give you an exact answer but of course the complexity there is the ISP in the middle. On a weekday night many of our customers can download at least a standard definition product without any problem in the world, and yet on a Sunday afternoon or a Sunday night they might see some congestion in the network because that is usually the peak hour of the week, and we might get a speed related complaint the following Monday. Most of that is because of the current copper-based network for those people not yet on the NBN, but as the NBN rolls out we are certainly seeing fewer people complain. I do not know if we have ever received a speed complaint from those on NBN versus those on legacy ADSL.

CHAIR: Have you ever mapped your complaints across the country?

Mr Lorken : All the time, absolutely. We are an ISP ourselves as well as being a content provider, so we do that. Obviously, as an ISP, we want to make sure that all parts of the country are up according to the SLAs that we have with our Telco resellers.

CHAIR: So you could look at a suburb that you know has fibre to the home and map complaints compared to a fibre-to-the-node suburb.

Mr Lorken : At moment there is not really a big enough sample size, just because of where the NBN is at the moment.

CHAIR: It is just not there yet.

Mr Lorken : Certainly we intend to do that.

CHAIR: If we come back in a year and ask you that question—

Mr Lorken : Yes.

CHAIR: I want to ask you about last-mile congestion, Mr Sneesby. You may observe that I am fairly new to this portfolio, so please be gentle with me. How does that work? Why does that happen?

Mr Sneesby : When you look at how our product is delivered to the consumer, we use a content delivery network, as all other players in this space do. A content delivery network places servers—computers—out there in locations which are as close as possible to the population. All of our content gets pushed out to a physical location and server that is closer. The closer you get to the consumer, the shorter the hop is when you log into your device and start streaming your show, and the shorter the distance is over the internet that that stream needs to travel.

So for the most part our content is sitting out there close to our consumers and the population. When I refer to last-mile congestion, it ready is the major or the biggest inhibitor that we see in delivering a quality service. It usually happens in that last mile—in the case of DSL, the hop from the exchange—the DSLAM—to the consumer's household.

We trace the performance of our network on an ongoing basis, and the data tells us that the majority of those challenges in streaming come in the last mile, which is effectively provided by the retail ISP. Obviously we track that over time and we track it to a certain degree geographically. As I mentioned before, the data tells us that during those peak times in the evening we see overall speed reduction, when we look at our entire base, and we see at the same time a higher incidence of consumers who are either experiencing pausing in their videos, which is due to the video not being able to get enough bandwidth to get down to the consumer, or our adaptive bit rate technology is kicking in, which automatically reduces the quality of your stream in order to deliver a consistent streaming experience, so you are seeing a lower quality video stream on your television set.

CHAIR: The other question I would like to ask—this may be an impossible question to answer—is, in relation to the compression work that you are doing to make things faster from your end, is there a view about how far that can go? Are we there yet, or can we compress even more? Given that I do not understand the technology and I do not want to—

Mr Sneesby : I will try to simplify it. There are lots of different ways that the smart people out there are looking at to deliver a better quality stream in a lower bandwidth environment. Generally speaking, the standards that are being developed that are looking to achieve that rely on having more processing power in the end device.

CHAIR: So I have to buy a smarter thing?

Mr Sneesby : Exactly. As our devices—our phones, our tablets and our smart TVs—get more processing power, they become more and more able to compensate using those technologies at the other end. So there is a dependency on the creation of that technology, which is that the new standards that I referred to before are fairly well advanced. There will still be a dependency on this turnover of consumer devices to catch up and be able to use that.

If you were to roll out that technology today, you would find that more recent devices, which have that greater processing power and the technology built in, will be able to accept a much lower bit rate and compression, but older devices will not be able to. So it is a combination of the smart technology at our end, but it has to be enabled by faster and faster processing on consumer devices as well.

Mr Meagher : That does throw up public policy issues about equity of access and those sorts of things. You will recall that during the transition from analog to digital television there was a big issue with legacy devices, and sometimes those legacy devices were in the homes of people who were perhaps less wealthy and could not afford to keep upgrading those things. The truth of the matter is that the technology will continue to improve, undoubtedly, but it is not a smooth progression. We have to be conscious of that.

CHAIR: From a public policy perspective also, we cannot rely on the consumer having to bear the cost of allowing that equation to work.

Mr Lorken : I think it is fair to say that, to date, that has been hidden in the prices of the next generation of iPhones and smart TVs and what not. Once you divide by the number of consumers, it is almost an immaterial cost. That may not always be the case, but we do not see it on the horizon, at least at the moment. Even with the next generation that Stan was talking about, which is called the high-efficiency video coding—HEVC—which is already out in the US and what not, you do not see an uplift in the prices of TVs or devices at the moment.

CHAIR: Can you adapt them?

Mr Sneesby : You cannot adapt an existing one. One of the things that generally benefits the speed at which the technology can be taken up is that we are seeing a shortening of the technology consumer life cycle. The turnover of devices generally, whether it is a tablet, a phone or a television set is continuing to quicken as the price of consumer electronics gets cheaper and cheaper.

CHAIR: It is amazing.

Mr Lorken : I will be a little bit presumptuous and push on to your question there. It also goes to how we in our industry pay a part in bandwidth. There are quite a few tens of millions of dollars a year going into compression and making sure that we are at the latest of technology. Of course, we are incentivised to do that, because we need our ISPs and RSPs to understand that our content is as compressed and encoded as possible. No party here would want any ISP de-prioritising any of our content because we are not playing properly in the greater game, if you know what I mean. Sometimes we are accused of taking free rides on the back of the networks, but certainly this is one example where it costs quite a bit of money to keep this stuff up to date.

CHAIR: Thank you. That helps me understand a lot better. Does deprioritisation occur now, or could it occur now?

Mr Lorken : It could occur now, especially for those DSLAMs that are congested on the older style network.

CHAIR: That is another piece of work that I have to understand. Is there anything else that you would like to put in front of the committee before we finish today?

Mr Lorken : There is one thing on the policy point of view. I think that nbn co is starting to look at this, although we have not had confirmation. Certainly one of their stated goals is to have as many active users on the NBN as possible. We certainly all concur with that. One of the ways for the end consumer to get onto the NBN is through an RSP that happens to be a wholesaler, which then goes to an RSP directly to nbn co. Foxtel broadband is an example, where we might use Telstra as a way to aggregate or find a way to the NBN network. I do not think that nbn co has necessarily thought through that. There are still high level capital costs to connect to the NBN, if we were to go to a full national network. Certainly, we see that one favourable avenue is to go through someone like Telstra or another aggregator in order to get there. We would like to have assurances from nbn that if we pay for a particular bandwidth level, that is able to be flowed through to our end customers. For example, if we pay for a certain level of bandwidth through the CVC pricing for, let's say, one megabit per second, because we are a content provider as well, it might be that if we go through Telstra—just as an example—their entire base may only need 500 kilobits per second. How does nbn ensure that our end customers are being provided with the bandwidth that we have paid for?

CHAIR: And they have paid for too.

Mr Lorken : And they are receiving revenue for.

CHAIR: Thank you very much. If you have anything further for us that you want to add, please do so. You did, Mr Lorken, take a question on notice. If you could get that—and any other information you would like to provide—to us by 31 March, that would be terrific. That concludes today's public hearing. I want to thank all the witnesses who appeared today, our Hansard and sound people and our secretariat.

Committee adjourned at 14:56