On Monday, July 8, 2002, at 10:06 , Ken Chang wrote:
>>> why should support a technology that has no market ... WAP?
>>
>> You better check up on the background before you make like that. PHS=20=
>> has a
>> huge market and is the main cordless technology and one of the major=20=
>> WLL
>> technologies on the planet.
>
> the success of data PHS is based on the failure of the voice PHS,
This is neither a failure of PHS technology nor a failure of the=20
operators.
It is not a failure of the technology, because PHS voice is very=20
successful outside of Japan, in fact wherever there is PHS, it provides=20=
the best service at the lowest price.
It is not a failure of the Japanese PHS operators, because they were=20
being promised by the Japanese government that the transit cost for=20
carrying their calls from one location to another would be drastically=20=
reduced and the government did not fulfill that promise.
> that the operators made excessive investment without a business
> case.
Nonsense.
Taiwanese, Chinese and even the Japanese PHS operators, everyone of them=20=
had a sound business case.
However, the Japanese PHS operators' business cases were initially based=20=
on continuing promises by the Japanese government to bring the cost of=20=
transit down.
How can you make a profit on a 10 yen / minute government set tariff=20
when you pay 10.5 yen / min for the transit ?
If you have government assurances that this cost will be drastically=20
reduced -and I am sure that dates and numbers were communicated of which=20=
the public don't know- then you do have a business case. Otherwise=20
nobody would have put up the money for building the networks in the=20
first place. Well NTT would have, because they are the ones who get to=20=
sell the transit. In a 1993/1994 context, how likely are the investors=20=
in PHS to not trust the government's assurances ? Very easy now to blame=20=
them. Hindsight is 20/20.
In any event, this is not a result of the PHS technology nor its=20
philosophy. You had set out to argue that PHS was not and could not be=20=
competitive. This is clearly not the case.
Japan is the wrong example for reasons explained. PHS implementations in=20=
other countries have shown again and again, that PHS is the most=20
competitive technology out there to deliver high quality voice services=20=
both in urban and in rural areas.
> there are some examples in the PDC world, too. one is
> the success of i-mode that DoCoMo updated to PDC-P without any
> real market study, and the other is TU-KA messenger which is made
> possible because it got extra capacity by losing customers.
None of these examples are in any way related to the Japanese PHS=20
situation.
Originally there have been the following two business models for PHS in=20=
Japan:
1) The market-share-first profitability-later model
As was the case in any emerging cellular market, the new operators in a=20=
newly created market could focus on capturing market share initially and=20=
worry about becoming profitable once the battle for market share has=20
been fought. In the Japanese PHS context, this involved to trust the=20
government's assurances to reduce the cost of transit over time in order=20=
to allow profits to be made eventually.
2) The loss leader model
A wireline telco with a country wide and dense coverage of the last mile=20=
could run a PHS service as a loss leader and make money on transit=20
charges. The idea is to make more money on the transit charges than=20
loosing on PHS.
These were the early models. Astel and DDI went for #1 and NTT Personal=20=
went for #2.
However, the Japanese government didn't undertake to do anything about=20=
the cost of transit and therefore, model #1 looked more and more=20
worrisome. =46rom a strategic analysis point of view, it is absolutely=20=
clear that it would be DDI to be the first to abandon model #1. It could=20=
not possibly have been any of the others.
In case you don't understand why this is so, try to find a summary on=20
Von Clausewitz' works, which have become standards in marketing=20
strategy. In a typical Clausewitzian scenario, the early Japanese PHS=20
market can be described like this:
Astel was occupying the top of the hill, NTT Personal was besieging the=20=
hill and DDI was waiting in the woods nearby looking out for an=20
opportunity to take some territory the other two neglected while in=20
battle for the hilltop.
DDI then identified such an opportunity in wireless data, and henceforth=20=
focused on ...
3) The data service model
A PHS operator to give up on voice as a bread and butter revenue source=20=
and instead focus on data services.
Now the Clausewitzian battle scenario looks like this ...
Astel is still sitting on the hilltop, NTT Personal is still besieging=20=
it, while DDI have taken a village nearby. Astel has to pay heavily for=20=
sitting on the hilltop and NTT Personal is totally comfortable in the=20
situation they are in.
Two to three years later the situations has emerged into this ...
DDI have build their village into a flourishing town. Astel left the=20
hilltop and NTT Personal had no interest in taking it. As a result DDI=20=
took the hilltop, without taking any casualties and due to their nearby=20=
town on the foot of the hill they don't even have to fear any siege. The=20=
generals in charge of NTT Personal now ask themselves what the point of=20=
a siege is when you then do not take the hilltop after all. As a result=20=
both Astel and NTT Personal have to find new strategies in the new=20
battlefield. They are now trying to learn from DDI and are in search of=20=
some undisputed village they could take.
Two new models came out of this ...
4) The alternative data service model
A mobile telco could run a PHS subsidiary as a data service add-on to=20
their bread and butter mobile service. The idea is not to cannibalise=20
their own cellular market share while at the same time using PHS data=20
services to complement the cellular offering.
5) The bypass transit monopoly model
A wireline telco with it's own last mile infrastructure could run a PHS=20=
subsidiary as a mobile add-on to their bread and butter wireline=20
service. Using their own infrastructure where possible, they could avoid=20=
overpaying transit charges to other networks and thereby make money on=20=
PHS voice services. The idea is to reduce the cost of transit well below=20=
the government fixed PHS tariffs.
NTT Personal became integrated into NTT DoCoMo and model #4 was chosen=20=
for it. Doccimo is the name and from a corporate strategy point of view=20=
it is intended to be an interim.
Astel was sold to TTNet, an ISDN wireline operator, and model #5 was=20
chosen for it. Astel's focus is no longer public PHS, but VHE for=20
TTNet's wireline customers.
And here is how Clausewitz would have described it ...
DDI have now build a city state around the hilltop and are unlikely to=20=
be challenged by anyone as they expand even further. NTT Personal, now=20=
under a new king have taken a nearby town under the protection of a=20
foreign power. Astel have settled in a remote village hoping that no=20
hostile army will notice them there.
With hindsight, if any of the three had foreseen before launch that=20
transit cost wouldn't change for at least another 6 years and then only=20=
be reduced very slightly, at least one of them would probably have set=20=
out to build their own transit network and they would have taken the=20
lead in the PHS market with voice.
Was it foolish for the PHS operators to trust the government ? Perhaps=20=
yes, but who are we to judge them. If it had been you and me, in the=20
context of 1993/1994, we would have probably fallen into the same trap.
I have been working in many countries for new mobile operators in a=20
newly created mobile market. Very often, one of the first questions the=20=
business planners had to answer was "Do we trust the government and use=20=
their infrastructure or do we want to build our own transit network?".=20=
Thus, business planners/investors with experience of building networks=20=
in emerging markets, might have had that wisdom and decided to build=20
their own transit network, but
a) The Japanese didn't have that experience back in the early 90s and
b) Japan is no third world country, a Japanese investor is unlikely to=20=
mistrust the Japanese government.
So, the lesson here is not that PHS technology and concept doesn't have=20=
a business case. The lesson is that three Japanese competitors, being=20
Japanese, are likely to all go for the same mistakes, where in a foreign=20=
context three competitors may have taken three, but at least two=20
different routes of which at least one would have been a successful=20
formula.
> why DDI can provide faster speed and fixed monthly access fee
> is because their good good network will be wasted anyway.
>
> it=EDs an excellent service I agree, but I=EDm afraid that the window
> is now closing. there are so many Japanese companies, electronic
> manufacturers, are behind Wi-Fi now, and the competition will
> bring the price down very quickly, like the broad-band access.
Again, why are you so limited by your own perception? Widen your=20
horizon. PHS is not predominantly about data. Instead it is the ideal=20
technology to provide mobile telephony to densely populated countries.=20=
The only reason why it is now being "abused" as a wireless data solution=20=
in Japan is the prohibitive cost of transit and government fixed =
tariffs.
Take this away and you have a killer app in PHS voice.
Unfortunately, in the current climate, nobody is going to shell out the=20=
money to build an alternative transit network all over Japan, but as we=20=
have discussed, WLL, becoming more and more widespread, may eventually=20=
provide PHS operators with an alternative to NTT and allow them to make=20=
a profit on voice.
On the other hand, the Japanese government will have to eventually push=20=
for reform. An agreement reached with the US government last year has=20
secured obligations that NTT will have to slash their wholesale rates by=20=
40% over a number of years. (The figure escapes me right now). So we're=20=
getting there slowly.
If you then also take away government price fixing for PHS, you may=20
eventually see 2 or 3 yen per minute country wide calls on PHS, perhaps=20=
even lower than that. In China, a PHS call costs a fraction of 1 yen. I=20=
think it's 0.167 US cents per minute, but I may not remember correctly.
What do you think will happen to DoCoMo's PDC user base if PHS tariffs=20=
go well below 5 yen a minute ? Everybody would flock over and within two=20=
or three yeas you could easily shut PDC down. This would be a blessing=20=
for Japan, because it would allow the Japanese government to clean up=20
and harmonise the Japanese cellular radio spectrum.
KDDI could then be assigned spectrum in the proper CDMA band and become=20=
100% compatible with the rest of the world, reducing cost and improve=20
services. DoCoMo and J-Phone could migrate to either CDMA or GSM or =
both.
The initial capacity of the PHS network was 32 million users. Since then=20=
capacity has been enhanced. Even if further capacity was required, this=20=
could be done more cost efficient than with PDC. Yes, everybody would=20=
have to buy new phones, but that is happening all the time anyway and it=20=
would be fuelling the industry.
As you can see, PHS is not a dead end, but one of a few magic keys the=20=
Japanese government holds to reform and kickstart the domestic industry=20=
and contribute to a recovery of the Japanese economy.
regards
benjamin
Received on Tue Jul 9 07:57:31 2002