SMART Letter #74
Crisis and Revolution in Telecom
July 23, 2002

             SMART Letter #74 -- July 23, 2002
            Copyright 2002 by David S. Isenberg
   -- "danger and opportunity" -- -- 1-888-isen-com

>  Quote of Note: Jeff Bezos on two kinds of retailers
>  Crisis and Revolution in Telecom by David S. Isenberg
>  Quotes of Note:
>     Gordon Moore on missing the PC
>     Steve Ballmer on how "Linux changed our game"
>     Peter Seebach on usability and video games
      Sir Arthur C. Clarke on the environment
>  Conferences on my Calendar
>  Copyright Notice, Administrivia

  "There are two types of retailers: those that work hard 
   to raise prices and those that work hard to lower 
Jeff Bezos, CEO, quoted in Financial Times, July 
22, 2002, p. 1.  

by David S. Isenberg

[This is adapted from my Communications Keynote at the 
World Technology Network Summit, New York, July 21, 2002.]

There's an "utter crisis" in telecommunications, says 
Michael Powell, Chairman of the U.S. Federal Communications 
Commission.  WorldCom, Qwest, Global Crossing, Adelphia, 
Enron, Deutsche Telecom, France Telecom; yesterday's Wall 
Street winners seem to be leading the entire economy into 
the abyss.  

Let's try to see the current crisis in perspective; it is 
but one phase of the Communications Revolution itself.  It 
has been said that a revolution is not a dinner party.  
There are crises.  Outcomes are ambiguous.  The powerful 
struggle to hold onto power.  The forces precipitating the 
revolution struggle to advance.  There are losers.  It is 
not pretty.  The final result is not knowable.  

The drivers of the Communications Revolution overlap, but 
are not identical to the drivers of the current crisis.  
The drivers of the current crisis are: 
   1) Overcapacity, 
   2) Bad debt, directly driven by technological advances, 
   3) A wholesale change of business models more profound 
      than the shift from horse and buggy to the jet age.  

I am tempted to add incompetence, greed and criminal 
behavior, but these are constants that are neither unique 
to the Communications Revolution nor the current crisis.  

With the benefit of hindsight (and hindsight is important 
if we are to learn from history), we can appreciate the 
power of the twin factors of overcapacity and bad debt.  I 
saw the third factor in 1997, the coming shift in business 
models, and I wrote about it in a paper called "The Rise of 
the Stupid Network".  This shift is still not a broadly 
acknowledged factor in the current collapse, but I think 
that it will prove to be the biggest factor.

Let's look at each factor individually to see how it 
relates to the current crisis and to the larger 
Communications Revolution.

Let's not call the current overcapacity situation a 
"bandwidth glut."  Gluttony is one of the seven deadly 
sins.  The scarcity folks -- the telephone companies (and 
others) whose business is based on the fact that 
communications capacity is scarce, therefore expensive -- 
are controlling this "glut" dialog.  Nobody talks about a 
glut of clean air or a glut of traffic-jam-free roads.  No 
-- to an end user it is great to have a lot of cheap 
network capacity.

I would say, "Glut is good," but slogans like this seem to 
convey the wrong message in the longer term.  So let's just 
talk about overcapacity.

There's a supply side and a demand side to the overcapacity 

On the demand side, telecom growth used to be predictable, 
at about 5% a year.  The Internet toppled this assumption.  

Then for several years people thought that Internet demand 
was doubling every 100 days.  This finding was based on 
early data -- it turned out to be an over-estimate.  Today 
people estimate that the Internet is growing at around 100% 
per year.  

One hundred percent per year is still an outrageous growth 
rate.   If you start with one megabyte at year zero, 100% 
for ten years would yield 1000 megabytes at year ten.  But 
in contrast, 100% every 100 days, the former supposed growth
rate, would yield 68 billion megabytes by Year Ten.  So we 
overestimated the growth of the Internet by a mere factor 
of 68 million.

The effects of this over-estimate on capacity build-outs 
and revenues were amplified by telecom competition.  In the 
late 1990s the United States, and then the World Trade 
Organization, passed laws and made agreements to help new 
entrants break into established telecom markets.  The new 
entrants saw one trillion dollars of annual worldwide 
telecom revenue, and they assumed that traffic was doubling 
every 100 days.  They wanted a piece of this pie -- a pie 
that they thought soon would be larger than the entire 
global economic product.

Again, with the benefit of hindsight, there might have 
been some things that would have increased traffic enough 
to create a viable market that would support multiple 

For example, traffic might have grown faster than it 
actually did if the recording industry had not put the 
legal kibosh on Napster.  Some say that if it were legal to 
trade video files a la Napster, it would be so popular that 
we wouldn't have any overcapacity.  In fact, we'd have to 
install new long-haul capacity.  If this is right, the most 
effective short-term fix to the overcapacity situation 
would be to reform intellectual property laws to be more 
consistent with how people want to use the Internet.  

Also traffic might have grown fast enough to keep pace with 
installed capacity if AT&T, instead of throwing $120 
billion down the cable TV hole as they did in 1998, had 
used that money instead to build fiber to their customers' 
homes.  Such an initiative would have had enduring, 
intrinsic value.  And it would have laid the groundwork for 
continued United States leadership in communications.  

Everybody believes that fiber to the home is the end game 
of the Communications Revolution.  It is not expensive, 
about US$600 to $3000 per home with today's technology (and 
less in the future, and less with economies of massive 
scale).  But just as Qwest's 1997 transcontinental fiber 
build-out fatally maimed domestic long-distance (including 
Qwest itself), fiber to the home would kill the Incumbent 
Local Exchange Companies.  

Therefore, fiber to the home is not coming until the 
Incumbent Local Exchange Companies become considerably 
weaker.  Fiber to the home, in the context of the Internet, 
puts too much bandwidth and too much control into the hands 
of the end user for a Local Exchange Company to profit 
under any of the current scarcity-based business models.  
But some day it is inevitable that we'll have fiber to the 

Anyhow, the new competitors and the older incumbent 
telephone companies miscalculated how fast demand for their 
capacity was growing.  More importantly, they missed the 
fact that telecom prices -- and profit margins -- were 

On the supply side, the rush of many new competitors into 
the marketplace, all of whom wanted to attract customers 
with ever lower prices and many of whom built their own 
networks, helped cause this collapse, but there were even 
more fundamental factors afoot.

Moore's Law, the doubling of computer power every eighteen 
months or so, suddenly made it possible to run the most 
sophisticated communications protocols on very tiny 
devices.  Today we can put a router on a chip that is more 
powerful than AT&T's big international switches that 
require an entire city-block-sized room in a skyscraper in 
downtown New York.  Plus, our ability to jam bits down 
glass fibers or copper wires or through the air grew faster 
than Moore's Law.  Suddenly, telecom infrastructure wasn't 
expensive any more.

Bad Debt.
This brings us to the subject of bad debt.  The debt 
picture is tightly intertwined with the rapidity of 
technological change.  

Here is an analogy:  We've all had the experience of going 
into a computer store and saying, "I could buy last year's 
computer for US$500 or I could buy the whizziest new box 
for $2000."  The problem is that in both cases, our 
purchase would rapidly become obsolete -- next year's 
machine will blow this year's machine away.  We know this, 
but we can't buy next year's machine -- we have to decide 
on last year's computer, this year's computer, or no 

Telecommunications executives are faced with the same 
problem, except that the time intervals are around 30 
years, capital expenditures are in the billions of dollars, 
and the core telecom business is at stake.  The telcos must 
build and upgrade their networks, but once they put their 
money on new technology, they're stuck with it for ten or 
twenty or thirty years.  

The telcos finance the new technology with thirty-year 
bonds.  But the technology becomes obsolete much faster 
than that.  

Let me give you two examples -- ATM (Asynchronous Transfer 
Mode) and SONET.  ATM was the greatest new technology in 
the early 1990s, but now, less than ten years after the big 
telcos installed it, some of them are taking ATM out of 
their network.  SONET and SDH were the bleeding edge in the 
mid-1990s.  SONET was very cool; you could cut the line and 
the network would self-restore in a few milliseconds.  But 
it isn't anymore.  For both ATM and SONET/SDH, the 
administrative overhead was too high (leading to high 
operating expenses compared to newer, simpler technology) 
and the protocols were not appropriate for Internet 

The bondholders that financed ATM and SONET now are finding 
that an obsolete asset secures their debt.  They're holding 
junk bonds that pay interest rates appropriate for asset-
backed securities.  

ATM and SONET are not the only technologies that are 
becoming obsolete even as they're being deployed.  There's 
DSL and MMDS and 3G and WAP and a whole lot more.  
Technology marches on.  And it is not as if Telecom 
executives made the wrong decisions -- mostly they made the 
best decisions they could at the time.

The debt movie is playing at the Global Crossing theatre 
and the WorldCom playhouse -- but soon it will be playing 
at a telephone company near you.  Verizon and SBC and 
BellSouth will not be immune.  The accounting tricks they 
used to use don't work so well all of a sudden, so they're 
likely to try some legal and regulatory tricks -- or even 
get new laws passed -- to fend off their day of reckoning.  
But watch out.  Even supposedly strong Incumbent Local 
Exchange Companies are at financial risk, because they're 
caught behind the very same 8-ball of rapid technological 

But the biggest thing happening to Telecommunications is 
not overcapacity, it is not about debt or technological 
change.  The biggest thing is that the entire telecom 
business model -- or maybe we should call it the operating 
model -- is about to undergo wholesale change.  We're 
leaving the Horse-and-Buggy era of telecom and entering the 
Jet Age.  

Goodbye blacksmith.  Goodbye livery stable.
Hello . . . to what?  We don't know yet.

The telephone company business model used to be based on 
vertical integration.  The network was a voice network.  
The wires were voice wires.  The switches were voice 
switches.  You can say the same for cable TV.  The cable 
system was specialized for broadcasting video 
entertainment.  These were special-purpose networks.

The Internet, in sharp contrast, is a general-purpose 
network.  It will carry anything.  The Internet does not 
care whether it is carrying voice or video or financial 
data or email or pictures.  

The Internet pushes the decision "What to carry," to its 
edges.  It pushes the decision "How to use the network," 
right into the lap of the end user.  This is a direct 
consequence of the Internet's architecture.  

The Internet's job is internetworking.  That is, the 
Internet is a network of networks -- the Internet Protocol 
is designed to span the various component networks and to 
ignore the network specific details.  The Internet Protocol 
ignores even those network-specific details that add value 
to a given component network.  

So if the owner of a component network that forms a piece 
of the Internet tries to add value to his or her particular 
network, that value may be useful for a network-specific 
application -- such as telephony or TV -- but it is 
irrelevant for Internet-level connectivity.  The only place 
you can add value in an Internet world is at the edges.  

This means that in an Internet world, a network owner has 
no special advantage in adding value to their network, say, 
over somebody who owns a few servers at the edge and buys 

This single fact makes the telephone-company business model 
obsolete.  It also makes the Internet the huge success, the 
integral part of our lives that it is today.  

Think about all the killer applications of the last decade 
-- email, instant messaging, web browsing, streaming audio, 
ecommerce, Internet telephony -- you don't have to be a 
network owner to host these apps.  Indeed not a single one 
was brought to market by a telephone company or a cable 

A lot of new telephone companies entered the marketplace 
with the old business model.  Most of them went bankrupt 
trying to be little Bell Systems or little AT&Ts.  The 
vertical integration thing doesn't work these days.  I 
suspect that we will see the remaining big guys dying from 
the same disease, if financial problems don't kill them 

A word of caution.  Today everybody from George Bush to 
Mike Powell to the wise executives of Silicon Valley are 
talking about broadband, broadband, broadband.  But 
broadband without real internetworking, without the pure, 
stupid, end-to-end Internet, will be as useful as a 
television that can order pizza.  I'd rather have the 
Internet over a plain-old dial-up connection than broadband 
with some form of pseudo-internetworking.

So if you hear that somebody is going to "enhance" the 
Internet -- to make it more efficient, to Pay the 
Musicians, to Protect the Children, to thwart hackers, to 
enhance Homeland Security, to find Osama, or whatever -- 
this is almost certainly propaganda from the powerful 
businesses that are threatened by the Internet.  Remember 
that the Internet became the success it is today -- and the 
threat that it is to existing telcos -- because it is a 
Stupid Network, an end-to-end network.

So that's what's driving the current crisis in the 
Communications Revolution; overcapacity, debt, and a 
wholesale switch of business models.  

The Communications Revolution is not over.  In fact, we're 
still at the early beginning.

What's the next phase?  I don't know.  But people will need 
to communicate, and the technology keeps getting better, so 
there is no doubt that the future is bright.

If I had the new model for telecom figured out, I'd be 
implementing it, not writing about it.  Back in the 1980s, 
people first realized that Moore's Law would make 
transistors so plentiful that it would commoditize hardware 
-- and that software would be king.  They scratched their 
heads and asked themselves, "How do you make money from 
software?"  One guy figured it out.  

We're faced with the same kind of problem.  Who will be the 
Bill Gates of the new telecom?  So far, the front-runner 
looks to me like . . . Bill Gates (due to Microsoft Windows 
Messenger).  But maybe it is somebody reading this piece, 
or somebody we've never heard of.

That's likely to precipitate yet another crisis in the 
Communications Revolution.  

QUOTE OF NOTE: Gordon Moore
  "If you asked me in 1980, I would have missed the PC. I 
   didn't see much future for it . . . I thought 
   automobiles would be a bigger market (for 
   microprocessors). But the IBM PC kind of hit it off with 
   the public." 
Gordon Moore quoted in ZDNet News, July 10, 2002

QUOTE OF NOTE: Steve Ballmer
  "We [at Microsoft] have prided ourselves on always being 
   the cheapest guy on the block--we were going to be 
   higher volume and lower priced than anybody else out 
   there, whether it was Novell, Lotus or anybody else
    . . . One issue we have now, a unique competitor, is 
   Linux. We haven't figured out how to be lower priced 
   than Linux. For us as a company, we're going through a 
   whole new world of thinking." 
Ballmer: Linux Changed Our Game, by Rich Cirillo in 
VARBusiness, July 15, 2002

QUOTE OF NOTE: Peter Seebach

  "Video games demonstrate several important lessons about 
   streamlining repetitive tasks. One of the first is the 
   use of rational defaults.  Video games try very hard to 
   get the defaults right. In many turn-based war games, a 
   unit can either attack or rest at the end of its turn. 
   What's the default? To attack if there's an enemy unit 
   nearby -- otherwise, it will rest. The default is almost 
   always right. Many games simply favor a most-recently 
   used strategy for guessing at defaults. This simple 
   strategy is often a gigantic improvement over the user 
   interfaces of productivity software.
. . .
  "A lot of productivity software is nowhere near the level 
   of reliability that you'll find with video games.
   Productivity software manuals are full of warnings to 
   save your work frequently, because crashes can destroy 
   your work in progress. Video games offer saving as a 
   convenience to the user, who may want to go do something 
   else for a while. Crashes are considered unacceptable in 
   a video game; for some reason, though, with most 
   productivity software, they're simply a part of the 
. . .
  "Many video games are designed so that the user doesn't 
   need to be taught how to play; the designers assume that 
   the user will never read the manual.
. . . 
  "Most games allow at least some level of user control 
   over the interface. Most productivity software doesn't."

Everything I need to know about usability, I learned at the 
arcade, by Peter Seebach, on the website, June 


QUOTE OF NOTE: Sir Arthur C. Clarke
  "I hope we can clean up our environment." 
Sir Arthur C. Clarke, in response to a question about the 
quality of life over the next 100 years at the World 
Technology Network Summit, July 22, 2002.


September 17 or 18, 2002, Washington DC.  American 
Association for the Advancement of Science.  I'll be 
speaking to the AAAS interns and the larger Washington 
telecom policy community.  Tentative at this time -- write 
to me ( to see if this is actually going to 

October 8-10, 2002, Atlanta GA.  Fall VON.  I'll be giving 
an Industry Perspective talk at 10:45 AM on Thursday, 
October 10, 2002.  See

October 15-17, 2002, New Orleans LA.  Fiber to the Home 
Council Annual Conference.  I'll be giving a keynote (on why 
neither telco nor cable TV co will bring us fiber to the 
home).  Nothing on the website yet, but keep checking for information.

October 22, 2002, Boulder CO.  University of Colorado at 
Boulder.  I'll be speaking to Dale Hatfield's graduate 
telecom seminar and guests, 4:00 to 5:20 PM.  Contact for details.

November 7, 2002, New York.  Marconi Foundation Award 
Conference.  Tim Berners-Lee will get the Marconi Award.  
I'll be speaking about the infrastructure that makes the 
World Wide Web possible.  More details soon.

COPYRIGHT NOTICE: Redistribution of this document, or any 
part of it, is permitted for non-commercial purposes, 
provided that the two lines below are reproduced with it: 
Copyright 2002 by David S. Isenberg -- -- 1-888-isen-com 

[There are two ways to join the SMART List, which gets you
the SMART Letter by email, weeks before it goes up on the web site.  The PREFERRED METHOD is to click on and supply the info
as indicated.  The alternative method is to send a brief, 
PERSONAL statement to (put "SMART" in the 
Subject field) saying who you are, what you do, maybe who 
you work for, maybe how you see your work connecting to 
mine, and why you are interested in joining 
the SMART List.]

[to quit the SMART List, send a brief "unsubscribe" 
message to]

[for past SMART Letters, see]

[Policy on reader contributions: Write to me. I won't quote 
you without your explicitly stated permission. If you're 
writing to me for inclusion in the SMART Letter, *please*
say so. I'll probably edit your writing for brevity and
clarity. If you ask for anonymity, you'll get it. ]

David S. Isenberg            , inc.                         888-isen-com                       203-661-4798 
     -- The brains behind the Stupid Network --