With the impending rollout of the NBN Bundles which will be effectively forced on the whole industry by the end of October 2018, I feel it is appropriate to talk about the issues that the smaller RSPs and their users will be facing over the next 6 months.

In a nutshell as the rollout of the NBN nears completion we are seeing more and more users being forced to move over to the NBN. This has resulted in a margin squeeze (long predicted by industry giants such as Simon Hacket of Internode fame) between the higher wholesale prices of the NBN and the resistance by users to pay significantly more for their internet. Unfortunately this is a crisis brought on by the telco industry as a whole with its long obsession with selling products based on price alone. This in the past has lead to some extremely sharp practices, including deceiving customers into signing contracts for very poor quality products. The ACCC has had to step in and limit some of these excesses, which I applaud.

Unfortunately the smaller RSPs are, and will be for the foreseeable future, bearing the brunt of this squeeze. They typically have higher costs because they do not have the scale to connect directly to the NBN and instead have to go through an aggregate wholesaler (Vocus, Optus, TPG, Telstra or Wideband) who will want a slice of their margin. The smaller RSPs also have limited capacity to cross subsidise with either business traffic or with other products such as mobiles. These smaller providers know that if they raise their prices they will lose customers. If they are stuck, due to contracts, with buying a certain amount of bandwidth from their upstream wholesalers, then this revenue loss goes straight to the bottom line.

I am seeing retail prices being offered by the smaller RSPs that are completely unsustainable – without a good margin in there to pay for support staff, billing, data centre charges, etc., some I believe have to be running at a loss. We have seen some RSPs raise prices and/or cut the products they are offering and generally it hasn’t gone well. Speaking personally we have seen the growth in our residential business slow considerably by having to increase our prices.

My concern is that we will see many smaller RSPs go to the wall as the larger players with their significantly greater capital, lower fixed costs and ability to cross subsidise can ride out the storm a little bit longer before they too will have to raise their prices.

So what advice can I give to help ride out this storm? For RSPs it has to be to look after your customers, connect with them, tell them what you’re up to, find something other than price that makes you unique. Be open and transparent. Don’t treat your users like cattle to be herded around and discarded when no longer needed. Resist the urge to compete on price, because someone, somewhere will always find a way to do it cheaper than you. Concentrate on what makes you different from the bigger companies – be nimble, be pro-active, provide awesome service, have a great network.

For consumers I suggest you ask what is important to you about your RSP and the service they provide. Are you sure that it is worth changing provider because they have just announced a price rise or there is another provider out there doing apparently the same service for $5 cheaper a month. Ask your provider about their network, how much bandwidth they are allocating per client (2.5Mbps is great, 1Mbps is too little). Can they increase and reduce their bandwidth as they need to. If you see a cheap price, ask why. If there is no reasonable explanation it is probable that it is unsustainable and they will be getting into financial strife soon.

It is tempting to ask them what upstream provider they use (if any), however you need to be able to understand their answer. It actually depends on how they are buying their service. If it’s “Layer 3”, then they are acting like a travel agent for an airline, they have no control over the quality of their network at all. If it’s “Layer 2” then they are acting like an airline, leasing planes – they have much more control over the quality of the service but they have to be careful that don’t fly with empty seats. There is a final type, which includes Launtel, where they connect directly to (and buy from) NBN and just lease space on other provider’s fibre links. This type of RSP fully control their network, it is also the cheapest to run, but has significant setup costs and the RSP either has be pretty large or limit the areas in which they offer services.

So if you are looking for a new provider, ask some deeper questions, choose carefully (don’t do it completely on price) and whatever you do don’t get into a contract – there are just too many changes coming along – you may find your RSP has been bought out or connected to a service that has had it’s bandwidth constrained (performs poorly) to contain costs.

Damian Ivereigh
Launtel
16/08/2018

There has been a big three-way game going on in the Australian telco industry between NBN, the retail service providers (RSPs) and their users. Come Oct 31st 2018 the final act will play out and all the signs are that NBN will win.

The TL;DR for consumers is that the price of internet will likely be going up by at least 20%. However the good news for consumers is that the quality of their service should be maintained. We will not be returned to the highly congested days of 2017.

How did we get here?

NBN has always needed to get its ARPU (Average Revenue Per User) up to at least $52 to make the project financially viable. However it was struggling to get this as high as $43. The first cause was that many telcos were selling the cheaper nbn12 (12Mbps) service, particularly to less technical consumers who just wanted “an internet connection”. The second cause was that under the original pricing mechanism, telcos were allowed to skimp on paying for bandwidth (CVC) onto the NBN network. This also allowed them to keep the price down (the large telcos have always found it easiest to sell on price) at the cost of poor quality, congested, networks.

To make matters worse for NBN the RSPs were able to land the blame for the poorly performing network on NBN – consumers were only to happy to blame the FTTN technology for example. There was lots of politics and the ACCC announced it would start monitoring network performance of the major carriers. NBN needed to find a fix to both it’s financial issue and its political one. NBN needed to find a way to force the RSPs to pay for more bandwidth.

The game begins

So at around this time (Oct 2017), NBN announced they wanted to encourage RSPs to move people up to the little used nbn50 (50Mbps) speed tier. A program they called “Focus on 50”. They agreed to discount the nbn50 to the same price as the nbn25 until May 2018, however, and this was the clever move, they also agreed to boost the bandwidth (CVC) they were selling to the RSPs by 50% at no extra charge.

Miraculously all the congestion problems went away. The ACCC announced that after doing some measuring there didn’t seem to be much of a problem after all. The nbn50 now became the standard speed that most people bought and even those that did not, they benefited from the extra bandwidth being provided. NBN knew that once consumers got used to all the extra bandwidth they would not give it up easily.

At the same time as starting “Focus on 50”, NBN also announced the new optional “Bundled” speed tiers. They were “bundled” in that instead of buying a connection (AVC) and bandwidth (CVC) separately, they were purchased together. They included into the nbn50 tier 2Mb of bandwidth – about double the bandwidth that most of the carriers had been purchasing before Focus on 50, but at a price that meant it was significantly cheaper than buying the two separately. However the RSP is forced into buying that bandwidth. Then to totally sweeten the deal, they extended the Focus on 50 discount to Oct 31 and added double the amount of bandwidth included with the bundles (to encourage RSPs to move over) until the same date.

In another clever move by NBN, they effectively forced RSPs to have to choose between the old pricing and the new bundled pricing. An RSP could run both pricing, but it would mean splitting their user base which destroys much of the economics of different types of users cross subsidising each other.

Really only the nbn50 is a good deal on the bundles, all the other speed tiers end up being more expensive. The nbn12 and nbn25 bundles are the same price as the nbn50 bundle (thus making them more expensive) and the nbn100 and above bundles are significantly more expensive than their unbundled counterpart. However because typically 75% of an RSPs users are on the nbn50 speed tier, the incentive to move to the bundles is still strong.

However this will have a significant effect on the consumer pricing. I believe the days of an internet connection costing less than $100 a month will be over by Oct 31st. The prices of nbn100 and above will go up significantly (roughly $30/month for included 2.5Mb of bandwidth). There is nothing in theory stopping an RSP going back to the previous pricing, however it will also mean that they will return to the days of congestion that we saw in 2017. It is unlikely consumers will put up with this and will just agree to stump up the extra money.

Given the impending price rise, it would appear to make sense to get into a contract at the current price. However I would advise against this. While an RSP cannot change the price on you, they can constrain your bandwidth. They will not be able to afford to put you on the bundle plans, so will leave you on an old unbundled CVC with much less bandwidth. So you may find yourself stuck in a contract with a poorly performing internet connection.

What does this mean for Launtel and its users?

We have always provided significantly more bandwidth than most RSPs, so in a way this is just the other RSPs being forced to do what we have always done. Given that we already provide this extra bandwidth, we don’t expect to see any significant price rises for our users.

However I still don’t believe the 2Mb of included bandwidth is enough long term. We believe an RSP needs to provide at least 3Mb to provide a highly performing, no download limit network. This figure will only rise over time, leading to more congestion.

Under the bundle pricing further CVC bandwidth can be purchased and we will continue to do this. This means that we may be more expensive than the major providers but very much providing a superior product for those consumers that care about the quality of their internet.

Damian Ivereigh – CEO Launtel

Further reading:
ITNews – NBN Co to create a class war on its network
ITNews – NBN retail prices are starting to rise

Submission to ACCC DTCS 2018 consultation
Damian Ivereigh, CTO, Launtel Pty Ltd
12th April 2018

[NB:
DTCS – Domestic Transmission Capacity Service – “Backhaul”
LTIE – Long Term Interest of End users – “Consumers”]

I own and operate Launtel, a relatively small NBN service provider in Tasmania. Almost our entire market is in Tasmania so we are heavily affected by the pricing of both intra-state and across Bass Strait bandwidth.

I have two points I would like to make:-

---

1Gb limit now too low given current traffic
The current maximum of 1Gb, we believe, is now too low a limit for the DTCS to not apply. We are currently purchasing 2.5Gb and there is no way we could be considered a large provider able to negotiate good rates from the cable operators. In the NBN space it is not unreasonable to allocate 2Mb per connection – this puts 1Gb at just 500 customers.

---

“Wet section” uplift to Tasmania does not promote competition
We believe that currently the large uplift of pricing for the “wet section” across Bass Strait (while intended to promote competition in the cable operator market) has had the opposite effect in the retail internet market. We strongly believe the 2016 DTCS determination has failed the LTIE test for Tasmania.

No new players have entered or show any significant signs of entering the cable market

The cost of laying new cables across Bass Strait is extremely high, yet the bandwidth market in Tasmania is not large (approx 600Gb) – most of which is not supplied under the DTCS anyway (I.e in blocks larger than 1Gb). For a new entrant to lay a new cable to Tasmania, they would have to be confident that the market will grow to accommodate them and they would not just be taking business from the existing operators. Without a growth in overall bandwidth it would simply lead to higher overall costs (capital etc) being paid for by the same amount of bandwidth sold.

We believe this is such a massive barrier to entry that we are unlikely to see a new operator enter the market. In other words the “wet section” uplift does not, in practice, actually create more competition in the cable operator space.

National pricing leads to poor outcomes in the retail space for Tasmania

Most (if not all) national RSPs have standardised retail pricing across Australia. Generally the differences in backhaul costs to different places is considered minimal enough that they can just be spread across the RSPs entire business and the same retail price is charged. However, this backhaul cost is not true with Tasmania; where the costs are often many times higher. In this situation the RSP must either have a lower quality network in Tasmania (more congested or less redundancy), decide not to offer services to Tasmania, or deliberately run at a loss (i.e. Tasmanian internet connections are subsidised by mainland ones).

As a small RSP in Tasmania we are faced with this same choice, except that the last option of running at a loss is not available to us given that we have no mainland portion to provide for that cross subsidy. We are therefore left attempting to compete against companies that are either making a loss or at least running with much smaller margins. We have had to restrict our services to those willing to pay more or support a local business.

High prices leads to reduced bandwidth purchases

One of the biggest barriers to the overall reduction in price in bandwidth to Tasmania is the relatively low amount of bandwidth purchased. Given that much of the cable operators costs (or required returns) are fixed regardless of bandwidth sold (in other words the incremental costs per Mb are very low), anything that can be done to increase bandwidth purchased overall will improve the competitive situation. This has occurred repeatedly on the Mainland. Having a high DTCS price for Tasmania does the opposite.

Current DTCS pricing leads to poor LTIE outcomes in Tasmania

We contend that this extra cost of bandwidth across Bass Strait negatively impacts the LTIE either directly or indirectly. Directly by being sold a poorer quality network or inability to access some providers; and indirectly by having to pay more for their internet.

The soon to be implemented NBN 50 bundle is setting the benchmark for the standard residential grade NBN connection. It includes 2Mb of CVC and is being sold by NBN for $45/month. Let us assume that IP Transit or peering costs $10/Mb on average.

We note that the 2016 DTCS price for 500Mb from Hobart to Melbourne is $12,202/month ($24.40/Mb). The same sized “tail-end regional” connection is $2,762/month ($5.50/Mb).

The direct cost price of providing the NBN50 bundle to a regional area is: 45 + 2*(5.50+10) = $76/month

The direct cost price of providing the same bundle in Tasmania is: 45 + 2*(24.40+10) = $113.80/month

That is an increase of $37.80 – almost 50% – in the direct costs that someone (RSP or end user) has to pay per connection per month.

It should also be noted that we have been finding that, to offer an “unlimited” (no data cap) connection, becoming the standard on the mainland, requires in excess of 3Mb per customer – this widens the gap in the above calculation even further.

As bandwidth demand increases, Tasmania is in danger of becoming stranded as a high priced and/or poor quality internet region of Australia. This is particularly ironic given that is has the highest penetration of NBN FTTP and thus should have some of the best internet!

Evidence for our submission

Launtel opened residential services in October 2017, priced similar to the market. After rapid growth, by January it was clear the growth was not sustainable at the current prices. We had to reprice the 50/20 (the entry level internet) at $105/month for a 750GB download service – well above the market.

Aussie Broadband initially opened up services in Tasmania in September 2017, by April 2018 the whole of Tasmania is on “stop-sell” awaiting a bandwidth upgrade. This has been slated (according to their website) for 31st July – a significantly longer period than normal. Phil Britt (CEO) has written on Whirlpool that Telstra charge “drug money” for the link – https://forums.whirlpool.net.au/archive/2716701#r56143304 and they are expecting to negotiate for better pricing.
Phil Britt has stated that they have no plans to upgrade to do the same 10Gb links or provide redundant links that they are doing on the mainland due to their cost – i.e. they are providing a lower quality network.

Damian Ivereigh
CTO Launtel

Too many people (not our customers) complain that the NBN is too slow.

Well it’s not the NBN. The FTTN (fibre to the node) network, whilst crappy can still be good. The FTTP (pure fibre to the premise) network that most of Tasmania has is a beautiful network that’s reasonably reliable and flexible to fulfil data needs for decades to come.

So then why are some connections slow? The crew at CNet explain (sorry about the 15 second ad at the start).

Launtel Business network is a Zero Congestion network. Our Residential network is currently low congestion (much lower than our competitors), but will become a no congestion network in coming months.

If your speeds aren’t what you’re paying for, tell your current provider you want out of the agreement or just give our free 7 day trial a go! We can show you how the NBN was meant to be.

Currently there is a big debate raging in the US about net neutrality. This is the principle that carriers (ISPs) should treat all traffic equally. They shouldn’t be allowed to prioritise certain traffic or charge more or less for it’s carriage.

This is a big deal in the USA because they have a significantly poorer competitive landscape than we do due to allowing (even encouraging) “infrastructure competition”. This is where a company can build a network (fibre, HFC, wireless) and then sell services direct to retail on that infrastructure. There is no requirement for that company to sell access to that infrastructure to other competitors.

The net result is that most customers in the USA have access to typically one ISP on each of the technologies (ADSL, HFC, wireless or sometimes Fibre). If there is only one technology available, customers have no choice about which ISP to use. To make matters worse, just as in Australia, there has been a massive consolidation in the telco industry leaving just a small number of large players who have the ability (and have shown their willingness to) throw their weight around.

If you want an example of what happened when Australia tried infrastructure competition, you only have to look at the disaster that was the original HFC roll-out. Optus and Telstra chased each other down the street with their construction crews with the sole purpose of destroying each other’s business model. The net result is that neither made enough money to cover the cost and the roll-out halted.

In Australia we do not formally have net neutrality. Due to the competitive pressure between the number of ISPs at retail, there is probably little incentive for any of them to “behave badly” in this respect and therefore no great pressure to introduce regulation.

Having said that, Telstra has been known to manipulate traffic to their own ends, the obvious example being their unwillingness to perform peering with anyone. Telstra, uniquely, charge both their retail customers and the content providers (Facebook, Amazon, Google etc) for data to traverse their network. They can only do this because of their market power in Australia.

WiFi, the ubiquitous wireless technology that allows devices to connect to the internet over a local connection, isn’t performing as well as it used to.
While fixing the issue presents some significant challenges, most of them regulatory, the time is right to make a start.

When WiFi was first invented in early 1990’s (a key part of the invention was made by Australia’s CSIRO), we were not connecting many devices to the internet, indeed the internet itself was hardly used by the general population back then. Since then the growth in smart phones and tablets and increasingly the “internet of things”, all using WiFi,  has been phenomenal.

The attraction of WiFi is obvious – the ability to avoid having to run wires or cables to the device being used. Indeed many devices (particularly from Apple) really have no other way to connect to the internet other than WiFi. Couple that with the sheer growth in internet traffic (video being an obvious example) and the whole WiFi system is creaking at the seams.

So what’s the problem? The problem isn’t with the technology, but with the amount of spectrum that is allocated to it, how much interference there is and that there are no strong rules for use of this space . Spectrum in someways is like real-estate. Governments regulate it and essentially own it, leasing it out to the highest bidder (typically telcos or broadcasters). Spectrum is used for many different applications such as radio communication, TV, radio and most obviously mobile phones. The spectrum is measured in frequency (Hz) and as a rough rule of thumb, the higher the frequency the less distance it will travel. So the government leases out spectrum in physical areas, at the higher frequencies these areas are smaller.

However there are a couple of parts of the spectrum that were considered unusable for leasing out because they had too much interference from medical equipment, microwave ovens and the like. So the governments around the world left them free for anyone to use without a licence. These are the frequencies used by WiFi. So the part of the spectrum that many people use for the going about their daily work and entertainment is this relatively tiny area that the governments didn’t care about. It’s like the government took all the prime real estate, leased it out and left the rest of us with the rubbish dump.

To make matters worse the rules for using this spectrum are pretty lax and there’s nothing to stop one device from interfering with another. Unlike, say, the mobile phone system where the tower essentially controls who can transmit and when, to keep each phone from interfering with another. As an aside this is one of the reasons you shouldn’t use a mobile phone in an airplane – unlike on the ground – your signals travel much further and interfere with neighbouring areas (cells).

If you live in the city and do a scan around for the number of WiFi networks, it’s not uncommon to see hundreds of them. All these separate networks are all using the same spectrums where only one device can transmit at a time. It’s no wonder that the networks start slowing down as messages have to be retransmitted. It’s like trying to shout across a nightclub. To make matters worse, the tendency is to get more and more powerful transmitters (though the regulations limit their overall power) to try and break through this noise. This of course has a detrimental effect on everyone else.

I’ve seen businesses decide that for convenience (or to avoid the cost of cabling) they’re going to go entirely wireless and then wonder why despite their super-fast internet connection they’re seeing buffering and slowdowns when they try to make a Skype call.

So what can be done?

The only real fix is to get more spectrum allocated. This will require lobbying of governments and agreement with the tech industry, who will have to design new devices to make use of it. Hopefully they will also agree a tighter protocol for managing this spectrum such that otherwise interfering networks can negotiate with each other. There is a new standard call 802.11n, which uses a higher frequency (so can pass much more data), but unfortunately it also uses another section of the spectrum that no-one wanted – it also travels a shorter distance.

It’s time that a decent area of the spectrum (and lots of it) was allocated for everyone to use. This is an important area of our lives and commerce and we deserve better than just having the bits of the spectrum that no-one else wants.

In the meantime my recommendation is that you use WiFi sparingly. There is no question that it’s extremely convenient, so use it when you need to, but avoid using it when it is possible to cable it in. For example when you buy an internet enabled TV, make sure it has an ethernet socket and use it. Video is extremely heavy on a network and putting this over WiFi is asking for trouble.

Launtel Business customers can already ask our advice for their individual office setup. After all, we’re investing a lot in providing you a fast and reliable NBN service, the last thing we want is it slowing down because of crowded Wi-Fi!