Thursday, August 07, 2008

Why would Telecom NZ launch an 850 MHz UMTS network?

Mauricio posted this rumour earlier today that Telecom is looking at UMTS at 850 MHz on a nationwide basis. I have no idea if this is true or not but I thought I'd assess it in advance on the off-chance it is true.

In my experience over the last 15 years with Mobile network deployments there have been a bunch of hypotheses about what you can do to attract customers to your network. These have included products and services, video, social networking - all the 'next big things'. 

In the end, it does not matter about differentiating on features and services if you haven't got three basics at the same level as the rest of the market. In order, they are

1. Coverage (Where can I use it?)
2. Handsets / Devices (What devices are available?)
3. Price (How much does it cost? Can I afford it?)

This is what customers repeatedly have told me when I have asked them what is important. In fact, given the benefits of mobility customers surprisingly rate Price significantly lower than coverage or handsets.

Most of Telecom's strategic errors around mobile have been when they have forgotten about these 3 fundamental truths - in particular the first. I consider roaming as part of coverage - and Telecom has been on the back-foot here for about a decade. In particular, apart from a period recently when these elements were equal, most of Telecom's major losses in market share have occured when the people involved in setting strategy have been too worried about the economics and trying to engineer a short-term position. But I am off topic now.

Back to the rumour - Why would Telecom look at this particular path?

Marketing wise, they have been saying that they have the best nationwide data network (true - CDMA in NZ is still the best natiowide mobile data networ) and a global roaming proposition (true). The issue is that it's a very limited range of devices that can handle this and when it comes to data, a key part of mobile now, the global proposition is non-existent for UMTS. Worldmode phones, while great, only operate on GPRS as far as I know. And there is a very small selection - 3 at my count out of the portfolio of 24.

This position gets even worse with the current stated proposition for WCDMA. 3 urban cities at launch and  nationwide EDGE network. Not good enough. It reminds me of BellSouth's entry into the NZ market. Lots of marketing and a very inferior coverage proposition. Many years after Vodafone bought BellSouth and had pretty much equalised on coverage , I still had Telecom customers tell me they were still with Telecom because of the vastly superior coverage. Power of brand...

So a quick assessment of the speculated 850MHz nationwide network. Why would Telecom do this?

1. Coverage - given the current plan this would provide a massive boost for local data coverage, allowing Telecom to switch off its CDMA network earlier. It doesn't add anything significant for global roaming. The global roaming game is at 2100 MHz for WCDMA - end of story. Anyone trying to convince you of anything else is dreaming.

2. Handsets - This is the tough one to swallow. This strategy does not help the handset situation at all. It means a more specific spec for handsets which means less range to select from. My guess is that Telecom will be forced to run a reduced portfolio compared to Vodafone locally. All they can do is ensure they have a set of choices for each price-point . Shouldn't be an issue at the mid-range but I would anticipate issues at the low-end and the high-end. Telecom may try and offer low-end at CDMA. This would follow the same path tried for trying to maximise the value of the TDMA network. That didn't work. This won't either. Don't do it!!!

3. Price - Telecom will have a relatively empty WCDMA network. Where they could really do well is if they do what Vodafone didn't do when they launched their 3G network. Offer a really sharp data plan. 

So, if you are an investor in Telecom stocks consider the following

1. If Telecom announce an 850 MHz network AND can talk to a reasonable set of handsets tomorrow then this is very good news. Especially if the focus is coverage and handsets as opposed to video services etc. If CDMA closure is announced, even better. Means that the long-term picture for Telecom's mobile business is built on a reasonable foundation. 

2. If there is no word on 850 MHz network then it is too late for this to be available for launch - doesn't rule it out but means another year of pain in Mobile. Whether you hold or sell will depend on the commentary around mobile.

3. If the rumour is squashed and the focus is all on ecosytems of products and services then be very very worried. Those don't matter if you don't have coverage and handsets sorted. Value down the mobile component of the business and cash-out while you can.

If it was me, I'd do the following

1. Roll out 2100 UMTS as far as you can - aim to make this the core of the network. take the capital hit now and a long-term view on return

2. Backfill with 850 UMTS - necessary evil and provides a short-term coverage advantage if you go full nationwide. Or worst case equalises this position.

3. Aim to close down CDMA in 3 years. Most customers will swap out their handsets in this period anyway. Find ways to make this work without blowing cost of sale.

4. Introduce an all you can eat data plan that goes where Vodafone can't easily follow  - look for their profit pool and nuke it. Great for customers and great for the revenue line.


M Freitas said...

Hmm. Global roaming in WCDMA is 2100 MHz in Europe, but in reality it is 850 MHz in Australia (Telstra) and 850 MHz in the U.S.

Most new phones will be 2100/1900/850 MHz but not many will have 900 MHz (Vodafone's new network).

At the end Telecom gets the bigger roaming footprint with Australia and the U.S...

Miki Szikszai said...

Asia is 2100 MHz
Australia is 2100 MHz - more operators than Telstra. Think Optus, Vodafone and 3
US is all over the show - everyone will have a chance there at 1900 and 850.

Net net - technical advantage at 850 MHz however it will get undone at the handset level.

If I had a dollar for every time someone said the answer for a non-standard spectrum deployment was a multi-mode handset then I'd be a rich man. 850 MHz is non-standard - global weight of 900MHz deployments means that WILL be the global standard. Sure Qualcomm will say they have multi-mode chipsets (and they do) but the number of operators asking for this specific chipset combination is small. Basically Telstra and AT&T. That's the same two operators that drove TDMA into a niche standard.

What that means for Telecom is two handset portfolios

1. at 2100/1900/850 MHz for global roaming (think Worldphone II). That is going to be a smaller and more expensive selection than the more prevalent 2100 UMTS / 900 GSM combination

2. A more generic one for 2100 MHz - probably grey market - that will have limited coverage.

Either way I don't see this as a winning play - at the end the global standard will be 2100 / 900 - like it has been for a while

Anonymous said...

Only the US and Australia are looking at 850. Internationally there are more 900 networks being built around the world. Handset makers are rolling out 2100/900 devices before 2100/850. Telecom will be better off than it is but not as well placed as it could be for mobile devices.

Anonymous said...

Uh. there are 180 wcdma 850 devices. How will we survive?

Miki Szikszai said...

180 WCDMA 850 devices. Great if you want to stay in NZ, US, Mexico, Canada and Oz. What about the rest of the world? That's what customers are making their decisions on.

Neutralising Vodafone's advantage on handsets and coverage will require 850 /2100 at a minimum. There is no doubt there is a portfolio that can achieve this. A cursory glance at the AT&T website shows about 10% of their current range. Most of their 3G devices are 850/1900

Fact: there will be much less than 180 devices available for 850/2100 when Telecom launches - there were about 200 CDMA devices when we launched CDMA. Took a year and a half just to get the portfolio of devices from 3 - 10. Not saying it's impossible - there is a hell of a lot more to it than a standards body saying what the global portfolio of devices available is...

Juha said...

What's the difference in terms of subscriber numbers per cell for 850 versus 2100?

Miki Szikszai said...

In terms of capacity (i.e. numbers of calls / data throughput supported) - none, assuming the same channel bandwidth per WCDMA carrier.

850 will support greater coverage than 2100 - so less cells per square kilometre. Much better when you are launching a new network as you need less sites to provide initial coverage.

Anonymous said...

Agree with this

1. Coverage (Where can I use it?)
2. Handsets / Devices (What devices are available?)
3. Price (How much does it cost? Can I afford it?)

But there's something you missed. Customers often don't think it's important, but it is... it's called marketing and it needs to be about what's available... not what will one day be available... If customers don't get told to buy stuff, they dont!


Miki Szikszai said...


There's marketing and there's marketing - how about trying a conversation with customers as opposed to blasting them with stuff. the customer trial is a good idea to start doing this...


Anonymous said...

Why isn't TNZ putting Transmit filters in as a courtesy to the other players in the market?

The spectrum at that frequency is very crowded, and as a signatory to the ITU treaties on proper spectrum management, TNZ is obligated to make sure the adjecant spectrum is clear and available for other operators to use. This currently is not the case. Its like a rogue operator buying a chunk of spectrum and setting up a transmitter that effectively blocks the competition.

kimberly said...

All this information is very interesting. We know about The Network, how ever if we want to see a visualization can be result weird for every one. Although the image are complex and difficult to understand, it help us to visualize the social structures. costa rica investment opportunities is a network very interesting too.