(keitai-l) Re: The Gospel of 3G vs. non 3G [Was: 802.11 for voice]

From: Michael Turner <leap_at_gol.com>
Date: 08/13/01
Message-ID: <002401c123c3$ca820cc0$984ed8cb@leap>
This is way long, WAY long.  But Benjamin has a good
argument here, too good to be ignored.  I'll try to demolish
it, but I think I'm on shaky ground still, at the end.  You
be the judge.  (Just don't nod off, now, judge, y'hear? ;-)

[And just to avoid further misunderstandings, from here on,
my 'you' means 'Benjamin']

[me, privately, to CFB]
> >so much for Benjamin Kowarsch's "oh, just let
> >a buncha WLANs gel and we can kiss the
> >  mobile/wireless operators goodbye."

[Benjamin Kowarsch]
> Very funny you quote me saying something I never said - neither
> directly as quoted, nor indirectly as I suspect you may later claim
> to have meant.

Since it's not nearly in your style, I wouldn't worry about it
being seen as anything but the loosest paraphrase, at best.
I have some wiggle room with my "we" -- if you take it to mean
"people who spent most of their time in WLAN hotspots, and
who do most of their communication with other such
people."  I.e., people on this list, who almost certainly
match that demographic quite well.  (Cf. your use of
 'we' below, which imputes to 'us' some assumptions
that most of us never made, I'm sure.)

Basically, though, I wasn't holding myself to terribly
rigorous quoting standards for the simple reason that
I had written to CFB *privately*, with a deliberately
provocative statement, hoping to elicit more, even  if
it contradicted me (which it did, somewhat).   He,
however (perhaps unwittingly), replied to keitai-l.

Next time I'll add an "OFFLIST" tag.  For the quote
I was paraphrasing, see below.

But anyway, while we're on the subject....

> The whole point is that WLAN telephony has a good chance of becoming
> a successful *coexisting* low cost alternative, not despite but
> *because* of it's "flaws".

> Try to liberate yourself from a few assumptions we are all too
> willing to take for granted ...

Hey, now who's making assumptions about *me* :-/

> Assumption #1: Any alternative wireless phone system will have to
> wipe out any other system or it can only be a total failure -
> coexistence of cheap with cheerful is impossible.

A general (not ironclad) rule of marketing: there isn't much
space for players in anything that becomes successful.  Three
at most, excluding tiny niche markets.  Your idea probably
couldn't come in second, given its admitted limitations, and
third place is always precarious.

A general (not ironclad) rule of "discovered commons": if it's
got potential for widespread use, industry will go to government
for patents and regulations that provide profit guarantees.

Depending on the government, the quid pro quo might be to
require guarantees of a competitive market (or at least the
appearance thereof) from those companies.  But definitely
tax revenue, in any case.  Government will seek something
else as well: a way to monitor activity in the newly-opened
zone.  (Especially if it's near borders, as international airports
tend to be.)  To the extent that the government is democratic,
you'll then see citizen action for privacy guarantees as well.

I think you can see the costs and scale of organization
adding up already.

> The examples of PHS in China shows that there can be coexistence
> between two different mobile telephony services. An very cheap system
> with limited mobility (PHS cannot be used out of town in China)
> coexists with a more expensive system with high mobility (cellular).

That's fine, but in either case, you have operators, right?  This isn't
the scenario you imply when you say [direct quote]:

   "But operating a bunch of loosely connected WLANs is always
   going to be a lot cheaper than operating cellular infrastructure, so
there
   will be a limit how far mobile operators can lower their tariffs to stay
   competitive."

Which I took to mean: "WLANs with no particular central organization
can take over the urban mobile arenas."  OK, that's an overstatement --
cf. mobile phones, you're mostly talking about a "quasi-mobile" (seated)
user base.

There are a lot of issues here, still.

Traffic management not least among them.  There has already been
a lot of VoIP hype by people who have never directly faced the
issues of telco network management while confidently predicting the
incumbents' doom.  Well, they didn't even get "co-existence".
(egg-on-face trophy winner: the now-bankrupt PSInet.)  And, as
CFB points out, "they'll come after you."  "They" being vested
interests for whom "cheaper" means "dangerous" and who have
a lot more clout with regulatory authorities than any (probably
balkanized) collection of upstarts.  Welcome to the vagaries of
anti-trust, which is particular wimpy in Japan.

> There is no reason why a VoIP over WLAN could not coexist with
> cellular as a low cost mobile option and thereby have an impact on
> cellular tariffs. The mobility trade-off is made up for by the lower
> cost. Depending on situation and user there would be uses for either
> service.

You've still got a chicken-egg problem -- getting enough people,
complying with enough standards, and making it convenient enough.

Yeah, yeah, I see the guy with the laptop at the airport who wants
to communicate with the home office cheaply.  But you've been
in airports -- can you imagine them *not* charging a premium
for WLAN access (esp. for voice purposes) there?  Furthermore,
airports are often quasi-governmental organizations, subject to
regulations and political pressures ranging from the civic to the
international.  Can you imagine them *not* coming under
gun from the telcos, local, national and international, even
if they wanted to offer free VoIP service (which they wouldn't).

I also *don't* see the guy at the airport if almost all this cheap
optical bandwidth gets used for videoconferencing instead.
I won't see him at the hotel either.  There goes your (high
margin) business market.

> Assumption #2: Mobile telephony *must* have automatic hand-over
> between cells while a call is in progress or nobody will want to use
> it.

Thanks for "liberating" me from this assumption.  I once was blind
but now I see. :-/

See my note on PHS, below.

> If you are sitting in a coffee shop and you have just ordered some
> coffee and cake, you know you will be there for a while - if at the
> same time, you can make a phone call there for a fraction of the cost
> of a call from your cell phone, then the chance is that you will make
> that cheap call although you know that it is going to cut off if you
> walk out there.

And -- tell me if I have the old joke right --  "if your grandmother
was a man, she'd be your grandfather."  How is it that you're sitting
in the cafe with the right equipment (not just your own, but the
WLAN too) unless you made an investment expecting to be
able to talk with a lot of people?

And who can put you in contact with most of those people?
And who owns THEIR copper and fiber and what's their price?
We all have our price.  (And fiber, especially, is distressed
merchandise these days.)

See below for who I suspect will end up managing those WLANs.

> If the calls are cheap enough, people are easily coerced into giving
> up a bit of mobility. That's what happened with PHS in the early
> years. People knew that the chance of a dropped call was high if they
> were moving around while a call was in progress. Many people
> therefore gave up using PHS in favour of cellular, but many others
> did accept that flaw in return for cheaper calls.

Having been corrected myself on this point: it wasn't just that
PHS dropped calls when you were moving -- you also had to
be moving pretty fast (e.g., train or car) for it to really matter.

> Who says that a viable service needs to be hi tech ? As long as it
> has a significant cost advantage most people do not  mind flaws.

This reasoning has its limits, though, and besides, quality is what
the customer says it is.  If I can't talk on it with people I want to
talk to, when and where I want to talk with them, you couldn't
pay me to carry it around.  It's got to be built into my mobile,
or I won't mess with it.  I doubt I'm alone in that.

> Assumption #3: Any mobile telephony service *must* have full
> geographic coverage.

I didn't need "liberating" from this assumption, Benjamin.  I was
here, watching PHS and keitai explode simultaneously, in the
rather limited area sof Kantou/Kansai.  And saw PHS survive.

On the other hand, Tokyo is a world unto itself in many ways,
and let's not forget those "immature market" subsidies, which
you haven't mentioned in your business strategy.

Finally, nobody needed to get some special hardware/software
configuration for their new PHS/keitai: it was already compatible
with the existing telephony.  Trivial barriers for us can be deal-killing
sales objections for the average consumer.

> Once upon a time people were used to having to find a public phone
> box in order to be able to make a call while out of the house/office.
> And still today there is quite a number of people out there who use
> pagers or no mobile communications tools at all and for whatever
> reason rely on public phones.

Yes, like the guy who's supposed to call me soon about work ;-)

> If there are a sufficient number of WLANs available in coffee shops,
> bars, hotel lounges, office buildings then a WiFi device, used for
> email and organizer tasks combined with a VoIP telephony application
> can be seen as a kind of portable public phone. Again, if it is
> significantly cheaper to make a phone call from that WiFi-PDA phone,
> not everybody, but many people would choose WLAN VoIP and not use the
> more expensive cell phone, even if it means the extra inconvenience
> of first having to check whether there is coverage.

"Sufficient number of WLANs" -- who's going to make this investment
and THEN manage it for unanticipated VoIP traffic (see further on)?
And how do you keep those strong-arm incumbents from taking
a piece and charging what this technology is costing them?

It's a future I'd buy into in a minute, if I could see how to get there
from here.

> The important thing is to have base stations in places where people
> spend a little while anyway and are not just passing through.

As has been pointed out ad nauseum elsewhere: in Tokyo, that
USD $5 for a cup of coffee was really the rent for a seat in
a convenient location in a booming economy.  So this is the
"quasi-mobile space" and there's proof that people will pay
for it (albeit less these days).  This is encouraging.

Starbucks Japan is being run by a cosmetics company (connection?
I dunno. Great place to touch up your makeup?)  So try to think
of a synergistic partnership with someone who's already got a
lot of real estate coverage.  Maybe your WLAN CoffeeNet could
be a sideline for a coffee/cafe-equipment distributor, replete with
counter-merchandising of the needed electronics?  ("10 coffees and
you get a discount on SIM cards."?)

But cafes might view people yakking on their phones as something
they want to be paid for, by the minute, since that yakking represents
opportunity cost (annoyed customers going somewhere else.)  So
there's the price-per-minute again, which might tend toward what
the market will bear: current cellular/PHS rates.  Maybe more,
in this context: surcharge as annoyance tax.

I'm trying to apply some marketing imagination to that question of
how to get widespread, fast exposure.  I.e., What you call
"thinking out of the box" (below).  But it really seems like
a stretch, with more obstacles than you mention.  And once
we get there, how do we know it'll still be cheap?  Which
was the whole point.

> Assumption #4: Any mobile telephony service can only be offered if
> the entire network is owned by one entity and administered as one
> entity.

I don't agree with this assumption anyway, but there's a good reason
why it *tends* to be true in practice: mobile telephony still requires
base-stations, which are an evolving, and high-maintenance technology,
and which require a huge investment in equipment, training, operations,
vehicles, accident insurance.....  Mobile operations tend toward single-
owner/operator models not because of some Cult of Bigness so
much as because, as things stand, you've got to be big to make a go
of it at all.  This is how increasing returns works.  Not dogma,
just economics.  Just so you know I don't take this as an axiom.

Take away needing base-stations everywhere, though, and you still
have expensive problems to solve.

Bootstrapping what you're talking beyond early-adopter/people-
willing-to-fuss-with-it is a very big project, and poses
many of the same management problems as regular telephony,
so I doubt it goes mainstream under any other auspices than
those of the incumbent operators.  So much for your "loosely
connected" network.

> On the Internet many companies offer services even though they do not
> control the last mile and their customers may use various different
> suppliers and methods to access these services each time they log on.
>
> Likewise, a VoIP service company could offer services, without owning
> any WLAN base station themselves. One day, a customer may use their
> services from within an airport lounge where the WLAN is operated by
> the airline and comes as a free service to frequent flyers. Another
> day, that same customer may use the very same services from a hotel
> lounge where the WLAN is operated by the hotel and access charges are
> put on the hotel bill. Yet another day, that customer may use the
> services from a coffee shop where the coffee shop operates the WLAN
> and time limited access comes as a throw in with each consumption,
> i.e one cappuccino gives you 15 minutes WLAN access time.

...Utterly neglecting that airport lounges, hotels and coffee shops
all run their own local communications monopolies, as long as you're
on their floors.  Why would they price access much more cheaply
than cellular/PHS?  The tendency, if anything, is to charge you
for the added convenience.  "Sorry, sir, your PHS/Cellular won't
work on the premises, but you can rent one of *our* WLAN-
connected phones from the front desk -- or, order it by room
service and don't forget to tip."

> Assumption #5: WLANs can only be used for telephony if their sole
> purpose is telephony.

Again, I never assumed this, so I don't need to be 'liberated' from it,
thanks.

> In the aforementioned examples each WLAN operator has a different
> motive to roll-out and operate a WLAN. The airline wants to improve
> their frequent flyer program, the hotel wants to improve their hotel
> service and the coffee shop wants to encourage people to visit and
> consume more coffee. None of them needs to have a viable telephony
> business case to roll out and operate their WLAN - in fact they need
> not and will not have any intention for any telephony service at all.
>
> Yet customers could use those WLANs to access VoIP services and the
> WLAN operators could become agents of the VoIP service company, like
> today news agents sell phone cards.

Utterly neglecting that many of the people who work in these businesses
are there because they have no particular technical skills.  They have a
a WLAN put in for one purpose, and (sidestepping the bootstrap problems
for the moment) it gets saturated by VoIP (driving away customers who
wanted smooth data access and/or peace and quiet).   Who steps in to
help them upgrade their LAN (and perhaps redesign the space) specifically
for this purpose?  Maybe a telco: not those groaty LAN technicians, but
people who have been selling and servicing electronic voice communications
for decades.  "Headaches?  Always, with new technology.  We know
that.  We'll solve the problem...."

But not with a loosely-connected network.

> Assumption #6: VoIP cannot work because of technical limitations ...

Never assumed this.  (Who WAS this "we" you were talking about,
anyway?)

> a) bandwidth on the Internet
>
> Who says that phone calls have to go all the way over the public
> Internet ? As long as the pipe from WLAN operators to the VoIP
> service's switching centre is big enough it doesn't matter. From the
> switching centre calls could be routed and delivered conventionally.

Key words: "big enough".  I.e., bootstrapping will go to the operators
that are already big.  This is no longer the "loosely connected WLANs" of
the direct quote above.  And it won't be PSInet, I guess.  Leaving....
oh my god, could its name start with either 'N' or 'D', or both?  Highly
centralized, more headcount than many nations' armies, with pension
liabilities to cover, and an executive suite that's heaviliy intermarried
into the more powerful government ministries?  My white flag is
already attached to the stick.....

> My buddy called me from London the other day and told me that he's
> got a calling card where he can call me here in Tokyo from London for
> about 4p (ca.6 yen) per minute. The quality of our 1 hour 20 mins
> call was excellent at all times. That calling card service may or may
> not use VoIP between switching centres, but the fact is that they use
> the incumbent telcos for the last mile on both ends. A VoIP service
> operator could offer the same kind of pricing using WLAN as a last
> mile either at the originating or the terminating end or both.

Yeah, and all they'll have to do to get big enough for bootstrap coverage
is borrow money from....oops.  Telecom debt markets are tapped out to
the end of the decade.  OK, then, they'll just IPO and....um.....VCs and
baby boomers aren't ponying up.  Especially after seeing a lot of their
money sink from sight in Ricochet.  Hell, you probably couldn't even
get much investment from the telco itself, if it were interested.

I guess we could start looking under the sofa cushions....

> Please, anyone on this list, don't try to tell me if you could call
> your buddies or parents back home for 6 yen per minute from your PDA
> or notebook while having a cup of coffee at Doutor's you wouldn't
> feel like going there more often and use that service.

Yes, but what do the signs on the wall  there say about disturbing
other people?

And anyway (not to be inanely repetitive) how to you get there
from here?

> b) IP address unknown by callers
>
> Who says that callers need to know the IP address of the phone they
> are calling ? Who says they even need to know that the called party
> is on VoIP ? As long as the VoIP service company makes this
> transparent to users it doesn't matter. There are various
> technologies around to accomplish this.

All of which have to be adopted by enough people, pretty
suddenly.

> c) called party not always online
>
[snip].....

> With WLAN, all I can see is ..
>
> - for anybody thinking out of the box there are quite a number of
> interesting possibilities

To which I can only say: there are always interesting possibilities,
the question are ones of market opportunities and execution; and
infrastructure costs are not all there is to TCO.

> - infrastructure is going to be rolled out anyway without any VoIP
> aspirations by the builders

Precisely what I meant by "gel".

You still have to pay the costs of management -- including the
cost incurred by writing off investment in previous generations
of technology, training in that technology, ad nauseum.  The people
doing current mobile telephony won't let go of this lightly.

> therefore it seems pretty obvious that there will be various
> companies using various business models to try and exploit this
> emerging infrastructure and not all but some of those are likely to
> come up with something viable and succeed.

And be bought by NTT, if in fact they aren't NTT already.

> Most interesting I find, however, the double standards with which
> technologies are appraised.
>
> On one side I am being urged to believe that *no matter what pricing
> model*, people are going to watch TV on their mobiles simply because
> of mobility - "it is there all the time" - and despite a trade-off in
> quality.
>
> On the other side, that same gospel also tries to convince me that
> public telephony would not be used over WLAN other than by a very few
> hi tech freaks even if it was to be *significantly cheaper* simply
> because of a trade-off in mobility.

As someone who's with you on the first point, while putting up
quite a wall of flak on the second, I would resent the implicit
sweeping generalization (except that I'm such a nice guy ;-)

[snip]

> .....I would be very surprised if a person
> capable of thinking out of the box would not find a significant level
> of bias in favour of 3G, being the latest gospel.

Hey, don't go turning political differences into religious
differences -- we've got enough to argue about just
with the politics!

There's "thinking out of the box," (rather a cliche at this point,
ironically)
and there's recognizing that we're still all in bigger boxes than we
can easily escape from.  Some of these are defined by institutions
and politics, and good engineering is the loser.  Others, however,
are defined by sheer economics and customer psychology, and
are therefore ineluctable.  Engineering that doesn't respond to these
realities might be good, but it will be just as dead as if it had been
wretchedly bad.  If you're thinking outside a box with just electronics
inside of it, about how much better those gadgets could do a certain
job if only the world were your canvas, you don't have a big enough
picture to offer plausible predictions.

A closing anecdote: a network engineer who knew little or nothing
about telephony management attended a panel discussion about
handling the huge and very spikey volume of logged telephony events.
He couldn't understand what they were worried about, and, in the
Q&A, asked "Why not just filter them?"  The hall erupted in
guffawing laughter, and the polite answer came back -- "our definition
of 'event' is 'either a billing opportunity or likely source of customer
complaints'.'"

Unless you can cut those people out of the picture in one fell
swoop, they *will* go after you.

And another one: Floyd Kvamme, of Kleiner Perkins, some 15
years back, was asked why he was funding crazy people with
crazy technology: "Oh, I'll take a technology risk," he replied.
"Just not a marketing risk."  Your idea seems so technologically
plausible I can almost taste it.  But it has 'marketing risk' written
all over it, in blood, at a time when the gilded silk for executive
golden parachutes is in excruciating short supply.  So I'm not
holding my breath.

I'm not trapped under any of the assumptions you want to
impute to me -- I just think I see some hidden, and crippling
(if not fatal), assumptions in your business plan.  If I had to bet,
I'd say early-adoption crests in five years, the mainstream
in ten, but very slow for now.  That's time to go get an
MBA with a concentration in marketing.

I suggest letting this discussion cool down until it...well,
until it "gels".  How about it?  "Liberation"?  The Sierra
Madre is a little hot this time of year, comrade.  Save
your ammunition for the rebel's spring offensive.  In 2005.

-michael turner
leap@gol.com



[ Need archives? How to unsubscribe? http://www.appelsiini.net/keitai-l/ ]
Received on Mon Aug 13 09:48:38 2001