The world's stupidest mobile phone business

Scott Wickstein has some follow up comments on my recent Samizdata piece on mobile phone technology. (There is now a lengthy further discussion in the comments section of that post, which some people may want to read, although it does get fairly geeky towards the end). In particular he points out the mess that Hutchison telecom appear to be making with their recently launched 3G network in Australia. I agree, but there are two separate issues here. One is the mess being made by Hutchison specifically, and the other is the problems with the 3G technology in general.

In order to understand all this, it is necessary to go back in time somewhat, and to look at the history of the Hutchison company. The telecoms business is all an offshoot of a the Hutchison Whampoa conglomorate belonging to Hong Kong tycool Li Ka-Shing. In the early 1990s, a lot of countries issued 2G mobile phone licences, and in most cases they issued more licences than there had been 1G (analogue licences). Hutchison bought new entrant licences in a lot of places, and built and operated networks under the name "Orange". In most instances (but not all) these networks became very successful and Hutchison sold many of them to other players at the height of the telco boom and made a huge amount of money.

When 3G licences were issued in the late 1990s, Hutchison attempted to do the same thing again. They bought 3G licences aimed at new entrants in a whole host of countries, and they are now attempting to build networks and market share as new entrants. However, it seems much harder this time, as the technology doesn't work yet, the incumbents have a lot more market power than was the case when 2G went into operation, and there are economies of scale that apply to 2G that do not apply to 3G yet. It is possible to provide new services using 3G such as video calls, but people don't appear to want this.

Now, however, we go to Hutchison in Australia. Here, the mismanagement is mind-boggling. To start out, there were only three 2G (GSM) licences issued in Australia in the early 1990s. These were awarded (on a tender basis rather than auction) to Telstra, Optus, and Vodafone. Hutchison did not get one. The next round of spectrum auctions in Australia were in 1998, when two licences were issued for spectrum in the 800MHz band formerly used by analogue services. Telstra and Hutchison bought these, and both built networks using the American technology known as CDMA. In many ways this is a superior technology (particularly in rural areas) but far fewer CDMA networks have be built in the world than GSM, and Nokia in particular is in the GSM camp. While Nokia makes CDMA handsets for the US market, it for a long time refused to supply them in Asian markets. This restricted Hutchison's product (again sold as "Orange") because the phones they had were mostly Korean and didn't appeal to the fashion conscious. Despite a lot of advertising, this, mixed in with the late launch compared to other operators led to Orange picking up very few customers in Australia. (They have an astonishingl lousy market share of about 2.8%. The market share otherwise is something like Telstra 45%, Optus 35%, Vodafone 17%).

For some reason, when more GSM licences went on sale in Australia in 2000, Hutchison spent something like $800 million Australian to buy one of these. Nobody knows why, as they have since done nothing with it. It just seems they pointlessly poured $800 million down the drain. (OneTel also bought a GSM licence at this point, and with the help of Lucent built a network, but this was switched off when OneTel later went bankrupt).

Thuse there are actually five 2G mobile phone networks in Australia. Telstra owns two (one CDMA and one GSM), Optus owns one (GSM), Vodafone one (GSM), and Hutchison one (one CDMA). (There was also one more (GSM) network that belonged to OneTel, but this was switched off when OneTel went bust).

Now we get to 3G. There are actually two separate technologies for 3G phones, UMTS (W-CDMA), which is European, and CDMA2000, which was invented by Qualcomm of the US. Existing CDMA networks can be upgraded to CDMA2000 without buying new spectrum, but UMTS requires new spectrum. (It is possible to build a CDMA2000 network in new spectrum, but not mandatory). If you have an existing CDMA network, then it seems to make sense to upgrade to CDMA2000, because the cost is a lot less. Also, there is the small fact that CDMA2000 actually works, whereas UMTS doesn't really yet.

In 2000, 3G licences went on sale in Australia, and as the number of people who wanted licences and the number of licences available were approximately the same, they didn't cost much. All the incumbents bought licences. At present, Hutchison are the only company to have built a network. However, they have built it using the UMTS technology that is incompatible with their existing network, and which doesn't work. (They have done this because Hutchison have globally decided to use the UMTS technology everywhere, and they are just obeying instructions from head office). Telstra on the other hand are upgrading their existing 2G CDMA network to CDMA2000 and at the same time have a licence to build a UMTS network if they discover later that this is the technology that is becoming dominant worldwide.

Thus we have a situation where Hutchison in Australia have built two incompatible mobile networks, have spent 800 million dollars on a licence for another incompatible network that they haven't built, and with all this they have managed to obtain less than 3% market share. This boggles the mind.

(And as for video calling being an application that nobody wants, I couldn't agree more. Picture messaging is clearly an application that people do want. (I am in Venice. Here is a photograph of the canal. You are stuck at work. Nah nah nah nah). People however generally do not want the other person to see them when they talk. There are two many issues of clothes, make-up, bad hair days, and the like. The stupid thing is that this was discovered by AT&T in the 1970s. AT&T actually wasted an enormous amount of money discovering it. Now, however, we simply have companies with no memory learning the lessons of the past all over again. The reason Hutchison have promoted video messaging as their killer application is that it is the only way they can really differentiate themselves from the competition. It is the only feature that they presently have that their competitors don't have. They are desperately clinging to it, because without it their whole business case collapses.

Oops: A half-written draft of this article escaped into the wild here. I shall finish it later. Since Scott Wickstein already left a comment, I will leave it up for now, but I will replace it with a finished version later.

Further Oops: It got partially updated, but then my computer crashed and I lost some further changes. This post is doomed.

Update: It's finished. Yay.