Is it Time to Rethink New Zealand's Approach to Mobile?

I’ve mused over writing this blog a number of times. I’ve alluded to the issues (I call them issues) a number of times in different forms, but what finally prompted me to put the energy into writing this was two very different views regarding our mobile market over the last month. 

The first was a conversation with a wholesale fibre provider who felt boxed out of the mobile market through regulation, and the second was after listening to an interview with a mobile provider who felt the market they were operating in was unsustainable due to demand driven by hyperscale tech companies (although somewhat at odds with their shareholders' report which would suggest improving margins?).

So both parties see a problem, albeit from a very different point of view. 

History

So first I would like to start with a bit of history. Who remembers Jetstream? In 1999 Telecom launched the first ADSL service in NZ under the name Jetstream. The Internet in NZ had moved beyond dial-up. However, all was not well. In the years that followed, Telecom continued to leverage its position as both a monopoly wholesaler and a retailer to the point that the government of the day gave them a stern word about changing “or else”. They chose the “or else” option, which saw a series of steps including local-loop unbundling and structural separation of Telecom. Communications Minister at the time, David Cunliffe, set NZ on a course to improve OECD rankings for broadband speeds.

This was closely followed by National’s 2008 election promise of an Ultra-Fast-Broadband initiative. The UFB initiative had a number of key tenants including:

• Structural separation - wholesalers (LFCs) and retailers (RSPs) were to be kept separate
• No overbuild - there was a desire to see limited overbuilding of existing networks. Essentially NZ’s population mass was too small to justify investment in multiple fibre networks, so overbuilding existing networks was discouraged, and LFCs were broken into regional areas
• Regulation - with a single network came the risk of a monopoly, therefore the focus was on creating a single regulated monopoly, thus preventing the potential for abuse of one's position

We ended up with a market with four UFB fibre providers - Chorus, TFF, Enable and Northpower, who had a regulated product set and price point, and could only wholesale these services to Retail Service Providers (RSPs). 

While UFB may not have been perfect (and in fact some of the regulations are arguably now restricting innovation and network investment), it has been a generational leap. Could you imagine working without fast, reliable and (mostly) affordable broadband? Throw in a global pandemic and the benefits are astoundingly clear. 

We can also compare the success of NZ’s UFB programme to Australia’s National Broadband Network (NBN), the latter being plagued most significantly with politics which saw flip-flops on budgets and delivery models and a long and expensive road to deliver less than UFB.

If you’re interested in more of the history you can read about it here: https://en.wikipedia.org/wiki/Internet_in_New_Zealand

Mobile

As I mentioned, UFB has been a generational change. It’s an investment that will continue to provide benefits for decades to come. Currently, more than 1.4 million homes and businesses are connected to fibre. 

In 2023 it was estimated that there were over 6.5 million mobile connections in NZ. Yes, approximately 125% of the population count and almost 5x the fibre connections (https://datareportal.com/reports/digital-2023-new-zealand).

Mobile connectivity currently dwarfs fibre connections, but that’s just the start. 5G has been designed to connect 100x more devices while providing ten times longer battery life for low-power massive-machine communications (https://www.mckinsey.com/featured-insights/mckinsey-explainers/what-is-5g). The World Economic Forum describes 5G as generational change  “5G is the technological answer, making possible billions of new connections, and making those connections secure and instantaneous. 5G will impact every industry – autos, healthcare, manufacturing and distribution, emergency services, just to name a few. “ (https://www.weforum.org/agenda/2018/01/the-world-is-about-to-become-even-more-interconnected-here-s-how/)

So if we thought UFB was a big deal, 5G and proceeding wireless/mobile technologies have the ability to be far more impactful. 

Spectrum

So that leads me to the topic of spectrum. For wireless networks to operate, they require spectrum. While your home wifi network operates on what is called the public spectrum, your mobile phone and many other services operate on the private spectrum such as broadcast radio. Private spectrum is licensed, with those licences being managed by the government.

It’s important that the spectrum is managed, as everybody trying to use the same spectrum creates interference. You may have experienced this in your own household using your wireless network, and having a cordless phone which ran on the same frequency. 

For mobile networks, most recently 4G, the way this has worked to date in NZ is that an auction has been held where parties bid on blocks of spectrum. The background to this can be found here: https://www.rsm.govt.nz/assets/Uploads/documents/8380f2d85e/spectrum-auction-design-in-new-zealand.pdf.

With the 4G auction Spark acquired four lots, Vodafone three lots and 2Degrees two lots (lots = 2x15MHz). 

Under the conditions of the auction, Vodafone and Telecom were required to build five and ten new towers each year respectively, for five years, in areas to which they do not currently provide coverage.

5G took an entirely different path. An auction for spectrum was not held citing COVID-19 “constraint” as the reason. Instead, spectrum was directly allocated to Spark, 2Degrees, DenseAir (a UK-based company that specialises in small cell connectivity) and the then Interim Maori Spectrum Commission. Vodafone was not allocated spectrum, seemingly on the basis that they had existing spectrum in the frequency being allocated (however they have subsequently acquired Dense Air’s allocation).

"As part of the new agreement, our three major mobile network operators - Spark, 2degrees and Vodafone - will be required to increase the pace of the 5G roll-out to small towns across New Zealand. There is also an expectation they will continue efforts to improve rural connectivity." 

- David Clark, Communications Minister at the time

There was also an allocation of spectrum to Tū Ātea (Māori Spectrum & Telecommunications Services). This does provide some scope for access to spectrum outside the major telcos. 

There was however no public spectrum allocation for private 5G networks like we have seen offshore (see Citizens Broadband Radio Service - CBRS). This is where there is now a huge amount of development of equipment for private 5G networks. One of the predominant use cases for this is IOT devices in dense environments such as factories. We are now seeing considerable development of lower-cost and easy-to-deploy 5G networks offshore. At a recent conference I attended, one of the biggest shifts was in the cost reduction and more compact size of private 5G infrastructure. 

Prior to the NZ 5G spectrum allocation, I raised my perceived need for a CBRS-like allocation with then Minister of Broadcasting, Communications and Digital Media, Kris Faafoi. His response was to highlight the Maori spectrum allocation, and that he felt they were “doing enough” - I paraphrase. 

The only thing I can find on the matter locally is a Radio Spectrum Management discussion document dated 2021 here: https://www.rsm.govt.nz/assets/Uploads/documents/consultations/2021-3-3ghz/3-3-ghz-consultation-document.pdf

Low Earth Orbit Satellite

LEO Satellite, popularised in NZ by Starlink, has had the most significant impact on the local telco market since UFB. All of a sudden rural broadband users had a real option for high-speed speed low(ish) latency broadband, with broad nationwide availability. The earliest adopters have been using Starlink for over three years now. It has become the network of last resort for residential use but also for a lot of industry applications including in our education sector. It has started to take a huge chunk out of Wireless ISP’s pockets, as well as mobile providers leveraging 4G infrastructure (which was often a very poor service). 

In a lot of ways, this is actually a positive for telcos. LEO services the less economically viable geographies, and actually helps therefore with overall margins. 

And Starlink is just the start… 2Degrees has started a satellite-to-cell trial with Lynk, and Amazon’s Project Kuiper is still to launch. These services become literal god-sends in natural disasters like the cyclone Gabrielle floods. You aren’t reliant on local infrastructure, with plenty of adaptations to run dishes off 12v feeds in a car (which becomes a pretty useful portable generator).

Not only has the arrival of LEO broadband addressed a major service issue, but Starlink recently made available a deprioritised plan for $79.95 a month - roughly half the price of the regular service, making it increasingly affordable too.

Mobile Black Spot Programme

The other thing that has never quite added up for me is the Mobile Black Spot Programme. We raise money through a spectrum auction (or don’t) and then cycle money back to the mobile providers to fill in coverage black spots, mostly around national highways and tourist locations. So ideally we want near-ubiquitous access, yet when we run our spectrum allocation auction we expect the opposite to occur. The providers here are acting naturally - they get spectrum, invest in assets and then seek to maximise the ROI. So, their investment is focused on the highest-density areas where they will get more uptake. 

This is most starkly observed with Fixed Wireless Access (FWA), where a consumer gets residential internet access via the mobile network with a mobile router, rather than leverage the fibre network which is often available in the area. Why? Simply because it cuts the wholesale price from the fibre provider out of the equation and increases the gross margin for the mobile provider, who can leverage more of their already invested in resources. A simple example may be a FWA service and a Fibre service may both retail for $80 per month, however, your ISP might pay the LFC (i.e. Chorus) something like $50-60 for the wholesale fibre service which their costs are all internal for FWA. 

One provider event took to naming their FWA product “FibreX”, presumably to compete with fibre offerings, which landed them a record $3.675M fine from the commerce commission.
Again, this is just smart business when looking at maximising your ROI but is it a good result if this is where spectrum assets are focused?

Looping back to coverage, and mobile black spots….if you want to find out for yourself if mobile black spots are still an issue, take a drive from Hamilton to Warkworth along some of our newest and most impressive stretches of SH1. Try to maintain a phone call for the journey and count the number of times your call drops. I recently saw a Facebook post where someone was unable to make an emergency call from Pine Valley Road, about 5 minutes from densely populated Silverdale and a main thoroughfare to the North-West of Auckland. 

Would a single mobile infrastructure provider result in a different outcome? Maybe.

The Point

I’ll come back to where I started - a mobile operator claiming that the market was unsustainable and they would need to begin charging more. The sustainability of the business model becomes a bigger problem if there is a need to broaden coverage into lower-density areas.

Also, I watch with interest the development of private 5G solutions offshore, with no clear path as to how they will make their move to NZ. AWS launched their private 5G solution in Aug 2022. You can simply order it, SIM cards and all, from the AWS portal.

Given NZ’s geography and lack of density outside major cities, mobile will be a key infrastructure component as fibre is impractical and uneconomic. It will support all sorts of use cases, in particular connecting devices, rather than people. With approximately 6.5M mobile connections in NZ, this number is set to soar, and with broader coverage (i.e. for primary industries use cases) and private 5G, this number could easily be 10X. Device SIMs need to be delivered at a much lower cost to a handset plan, they require less data, but they still need coverage.

So my questions are:

1. Is the mobile market sustainable for 3 providers to overbuild networks?
2. Is the UFB-style wholesale model more appropriate for shared infrastructure delivery of mobile?
3. Are our policy settings around spectrum strategically placed to allow us to take advantage of developments such as Private 5G or LEO direct to a cell?

The Rural Connectivity Group, a joint venture between Spark, 2Degrees and One NZ has demonstrated that cooperation can work to an extent, but why only leverage this model in such a narrow fashion?

Clearly, my opinion is that this needs a lot more consideration and that wireless-based technologies need a more strategic national infrastructure approach. It has worked well for fibre, so why not? Is it simply that we’re scared of upsetting the commercial apple-cart?

So what would I like to see happen?

• A CBRS-like allocation as possible and certainly immediate communication on the plan to get there
• Cost/benefit analysis undertaken on a shared infrastructure & wholesale/retail separation for mobile (with full transparency and open market consultation)
• A future looking strategic plan for the use of non-terrestrial networks in NZ
• If nothing else, it would be good to open up the debate and discussion on this topic.

When an eccentric US billionaire, with a constellation of satellites as a side gig, makes the most significant change to our telco market in a decade, the industry has moved on and we need to consider how we move with it.