SMART Letter #85
How to Monopolize the New Network
March 4, 2003
!@#$%^&*()!@#$%^&*()!@#$%^&*()!@#$%^&*()!@#$%^&*()!@#$%^&*()
------------------------------------------------------------
SMART Letter #85 -- March 4, 2003
Copyright 2003 by David S. Isenberg
isen.com - "turkey"
isen@isen.com -- http://isen.com/ -- 1-888-isen-com
------------------------------------------------------------
!@#$%^&*()!@#$%^&*()!@#$%^&*()!@#$%^&*()!@#$%^&*()!@#$%^&*()
CONTENTS
> How to Monopolize the New Network
> Quotes of Note
+ Dick Cheney on "Noise in the System"
+ Tom Ridge on "Noise in the System"
+ Paul Wolfowitz on "Noise in the System"
> If it's Funny, it Must Be True, by Scatt Oddams
> Conferences on my Calendar
> Copyright Notice, Administrivia
-------
How to Monopolize the New Network
by David S. Isenberg
What's the business context that brings the New Network?
Competition? Sorry, we can't get there from here, and even
if we could, we'd be up against Googin's Paradox (that the
best network is the hardest one to make money building and
running, see http://netparadox.com/netparadox.html). Is
regulation the best route to realizing our networked
potential? Nah. Stuck in the past, Byzantine, for
insiders only, not responsive to innovation in these times
of great change. So, maybe, hypothetically, pure hands-off
no-holds-barred unregulated monopoly, despite its ugliness,
will bring the most connectivity soonest.
The wise old marketplace has seen trouble -- exuberance,
oversupply and bust -- many times before. 1865, for
example, was a euphoric year in Oil Creek, Pennsylvania.
It had been six years since Edwin Drake triggered an
instant oil boom by showing the world that oil wells could
be drilled and oil could be pumped to the surface. By
1865, "the wild drive to produce created in the Oil Regions
[of Pennsylvania] a chaotic scene of heaving populations,
of shacks and quick-built wooden buildings . . . of
derricks and storage tanks, with everyone energized by hope
and rumor and the acrid scent of oil" [_The Prize: The
Epic Quest for Oil, Money & Power_ by Daniel Yergin, Simon
& Schuster, New York, 1991, p. 32-33.].
At the time, oil had one big application -- kerosene for
lighting. Kerosene gave better light than candles. It was
a lot cheaper than whale oil. In the twelve months after
Drake drilled the world's first oil well, some 15
refineries sprang up in Western Pennsylvania. "Kerosene of
widely varying quality was sold. If the kerosene contained
too much flammable gasoline or naptha, as sometimes
happened, the purchaser's attempt to light it could be his
last act on this earth" [ibid, p. 40].
The market for kerosene did not develop apace production.
"Toward the end of the 1860s, as overproduction caused
prices to plummet . . . the new industry went into a
depression. The reason was simple -- too many wells and
too much oil . . . between 1865 and 1870 the retail price
of kerosene fell by more than half . . . refining capacity
was three times greater than the market's needs . . . The
1870s were to be marked by ever-rising production.
Producers repeatedly tried to restrict production, but to
no avail. Storage tanks overflowed, covering the land in
black scum. The gluts became so large and prices fell so
low that crude oil was run out into streams and onto farms
because there was nowhere else to put it. At one point,
the price [which had been as high as $13.75 in the mid-
1860s] dropped to forty-eight cents a barrel." [ibid, p. 39-
42].
Standard Oil was so named because it brought standards to
the production of kerosene; Standard's kerosene would not
explode. Guided by the business vision of John D.
Rockefeller, Standard Oil created insider deals with
railroads to get rebates on its oil transportation costs.
Unbelievably, Standard Oil also negotiated "drawbacks,"
which the railroads paid to Standard Oil for oil transport
by non-Standard refineries. The control of oil transport
costs gave Rockefeller's Standard Oil an insurmountable
advantage.
Rockefeller wanted to end glut-and-bust. His aim was to
eliminate competition, "that cut-throat policy of making no
profits," and control production to "make the oil business
safe and profitable" [John D. Rockefeller's own words,
ibid, p. 42]. It took him a mere 14 years. By 1879,
Standard Oil controlled 90% of the refineries, and it
dominated oil production, railroads, pipelines, the supply
of oil barrels, and (some said) the legislatures of
Pennsylvania and New York.
The public was not delighted that Rockefeller's monopoly
sold safe kerosene and brought an end to the glut and-bust
era -- the Sherman Anti-Trust Act of 1890 was written in
direct response to Standard's "powerful, devious, cruel,
entrenched, all-pervasive and yet mysterious" monopoly
muscling [ibid, p. 96]; Standard itself was convicted under
Sherman in 1909 and broken up.
However, it cannot be denied that monopoly has an upside; a
robust and orderly oil business developed under the
Standard banner. Nor can it be argued that monopoly is
always against the interests of its customers; the first 60
years of the Bell System gave the United States the best
telephone system in the world. Nor can it be argued that a
marketplace with multiple competitors is free of criminally
anti-customer behavior.
If John D. Rockefeller were alive today, he would be
building fiber to the home (FTTH). Fiber has so much
capacity that once FTTH were built, there would never be a
need for a second fiber. The builder of the fiber could
tell the telephone company, "Hey, why bother to maintain
that expensive copper-based distribution network? I'll
carry those telephone signals to your customers for less
than it costs you to maintain that copper wire." Then the
owner of the fiber could tell the cablecos the same thing;
"Why bother with that coaxial cable? Just use my fiber."
Then the owner of the fiber could offer the same
proposition to Internet Service Providers, record labels,
movie studios and anybody else with traffic that a home
user might pay for.
If the first fiber builder worked single-mindedly to
aggregate all traffic, the way John D. Rockefeller worked
to consolidate his hold on oil production, oil refineries,
and oil transport, that fiber could form the basis for an
insurmountable monopoly.
The first builder of FTTH would have incredible market
power, even if its fiber only covered a neighborhood or a
few small towns. If a second fiber builder threatened, the
first builder could simply lower prices to ensure that
neither providers of downstream traffic nor end-user
customers would switch. When John D. Rockefeller did this,
he said he was giving nascent competition "a good sweating"
[ibid, p. 42]. A more recent monopolist referred to such
tactics as "cutting off [the competition's] air supply."
The builder of the first FTTH could offer end-user
connectivity that would not only be cheaper than a second
fiber, but would also be cheaper, more reliable and more
capacious than connectivity via twisted pair, coaxial
cable, or radio waves. But to do that, the builder of the
first fiber would need to be single-minded about co-opting
other networks and carrying all available traffic.
Like John D. Rockefeller, the owner of that first fiber
would have to pay strict attention to costs. The Internet
lowers costs because it is designed for reliable service
over unreliable media. As a result, fiber allows simpler,
lighter-weight protocols than telcos use. Fiber to the
home can be built and lit today for between $600 and $3000
per customer. Maintenance is another cost factor; in the
mid-1990s former Bell Atlantic CEO Ray Smith declared that
fiber was so reliable that it beat copper on truck rolls
alone, independently of any new business it might bring.
Scale is a third factor; once right-of-way is acquired,
ditches dug, et cetera, pulling 1000 fibers costs only a
few percent more than pulling two. Once a fiber builder
makes the decision to build at scale, there are solid
reasons to believe that costs could drop well below $600 a
home.
Given the FCC's February 20, 2003 order (written to update
FCC regulations enforcing the 1996 Telecom Act) that the
ILECs need not unbundle nor share their FTTH with
competitors, you might think that the telcos would rush to
build FTTH. You might think they'd rush to co-opt
complementary networks and carry additional traffic such as
TV signals. You'd think they'd rush to lock in future
monopoly access to the home. But they won't.
I've seen telco culture up close. The telcos are no
Standard Oil. The people running them are no John D.
Rockefeller. If some future John D. Rockefeller got a job
at a telco, he'd quit (or be fired). There's no room for
vision or visionaries in patrimonial telco culture.
No, unfortunately the telcos will not rush to lock in a
future monopoly. Unlike John D. Rockefeller, the CEOs of
telcoland are neither believers in, nor builders of the
future of their business. They want to see fiber optic
success first, before they build. Further, they'll never
expand laterally. They were not promoted for their
business acumen; they don't have the strategic ability to
co-opt competitors or neighboring businesses. And finally,
they'll want to build the old fashioned way, incrementally,
with constraints on capacity, with complex (so-called)
reliability mechanisms that will add to their costs.
In addition, the telcos are too worried about the present
to pay adequate attention to the future. Long-distance has
been in precipitous decline for years. Cellular service
revenues in the developed countries are maxing out. Now
the number of local loops is declining and local service
revenues are beginning to shrink as new ways to communicate
encroach on telephony-classic. Even ILEC data revenues are
hurting, despite the Internet-driven need for more DSL, T-1
and DS-3; according to Phil Jacobson and Farooq Hussain of
Network Conceptions, LLC (http://www.netconllc.com) each of
the ILECs recorded shrinking data revenues in at least one
quarter of 2002. This is not cyclicity, it is the
beginning the end of the vertical telephone company
business model.
It's even worse than that. The February 20, 2003 FCC order
perpetuates high-cost telco incrementalism. The FCC says
that it "elects not to unbundle bandwidth for the provision
of broadband services for loops where incumbent LECs deploy
fiber further into the neighborhood but short of the
customer's home (hybrid loops) . . ." Telecom analyst and
Former FCC staff member Kevin Werbach says that this
incents the ILECs to install "Potemkin fiber" -- a single
silly millimeter of fiber in the loop would protect the
ILEC from unbundling.
Telcos get this message; Scott Moritz at TheStreet.com
reports that Verizon CEO Ivan Seidenberg says that
Verizon's capital spending, which declined 30% in 2002 "may
contract" further in the wake of the February 20 order.
Instead of spurring a great new wave of telco investment,
the FCC's February 20 order will spur ever more steeply
declining telco revenues, a new wave of rear-guard
litigation and lobbying, and a growing gap between what the
technology makes possible and what the telcos actually
deliver.
There are several plausible incremental, non-monopolistic
scenarios for how the market for Internet connectivity
might evolve.
For example, my former AT&T boss, Greg Blonder, advances a
theorem in a provocative article in Barron's (Nov. 11,
2002, http://tinyurl.com/6hg0) to explain why today's
telcos are not rushing to embrace FTTH. He observes that,
"If new technologies promise improvements by a factor of
more than four, they tend not to get funded because they
are seen as too risky. And if they promise less than a
factor of two, they tend not to get funded because they
offer too little economic benefit." Fiber would have 30
million times the capacity of dial-up data when lit at the
highest speeds allowed by today's commercially available
technology. FTTH, lit at practical speeds, affordable for
a few tens or hundreds or thousands of dollars, would offer
100 to 10,000 times dial-up speed. No wonder the telcos
are not rushing to install FTTH.
Blonder observes that the hybrid fiber-coax network of the
cablecos provides a moderately fat pipe that is amenable to
a series of 2x to 4x improvements. He projects that within
a decade the current broadcast-based business model will
disappear because all video will be video-over-Internet.
From this, he reasons that cablecos are positioned to
migrate from a broadcast model to a connectivity model.
Supplementally, Blonder adds, cablecos could lay down a
substantial footprint of aggregated wireless ((e.g.,
802.11) hotspots. It is a plausible scenario, but I'm
skeptical that cablecos (or *any* company) can successfully
switch to a very different business model, especially when
the new business model is perceived by current management
as a threat to the old one.
There's another plausible scenario for the incremental
arrival of the next network -- primary radio connectivity.
Several people, most notably David P. Reed, have observed
that wireless networks can be architected so that their
capacity can grow with the number of customers. Economics
of such networks are quite attractive -- they require very
little advance capital expenditure. Put up a base station,
and after that you only need to spend build-money as
customers are added. Because capex is tied directly to
revenue growth, the network builder avoids the risky up-
front capex that Blonder's theorem takes as given. If
customers are willing to buy their own network
infrastructure, today's fixed-wireless technology can offer
up to 1000 times dial-up speeds.
Reed also addresses the scalability of wireless networks.
He proposes that the throughput and redundancy of a network
grows without obvious limit when network nodes can relay
packets that are addressed to other nodes. That is, every
additional customer brings more additional network
infrastructure than they themselves need. As Duncan
Davidson, co-founder of relay-network equipment builder
SkyPilot observes, "As a mesh gets dense, the links get
shorter, speeds increase, power drops, reliability
increases, line of sight goes away as an issue, and the
radios get simpler and cheaper." It reminds me of the
physical properties that drive Moore's Law. Reed calls it,
"Cooperation gain" [http://tinyurl.com/6sgm/].
In theory, wireless relay networks could be completely
user-owned; a service provider could be as free to add a
competing gateway to such a network as any other
participant (I dare not say customer). Networked nirvana
is at hand!
However, such concepts are still in research stages. Reed
notes that Claude Shannon's seminal work on information
theory was based on a single user, and that there is not
yet a complete multi-station information theory. First work
in that direction yields surprising and counter-intuitive
results; Reed points out that relay-hopping is not the only
thing that increases the capacity of a network. Multi-path
transmission and motion increase capacity too.
We are just beginning to explore packet relay systems.
There are many real-world unknowns. The first such multi-
hop network product, by Nokia Rooftop Networks, has not yet
taken the world by storm. Nor have we yet observed how
actual traffic behaves on such networks -- does it scale in
well-behaved ways or are there emergent, unpredictable
negative effects? In addition, we know that each station
requires non-trivial antennas, transmitters, signal
processing and computational logic that add to costs in
ways that fiber does not.
Even if these other scenarios were to become reality, an
FTTH monopoly could disrupt and subsume them. Surely,
somewhere on Earth, some budding monopolist, somebody who
is not immunized by interests vested in current networked
business models, could be attracted by the disruptive
potential of FTTH -- tens of thousands of times dial-up
capacity, or more, and growing. Once attracted, building a
business would still be a difficult, non-obvious
proposition. Getting the tactics right would be harder
still. Actually doing it would require even more
extraordinary business acumen, perhaps ruthlessly
discriminatory pricing, or collusion with existing
providers, or even willful disregard of the End-to-End
Principle.
It is likely that an FTTH monopoly would not be pretty.
It could lead to decades of criminally culpable networked
collusion. But t'ain't necessarily so. It might also be
the fastest way for humanity to benefit from the technology
we now have at hand.
Back in the early 1980s at Mattel, we didn't understand
Moore's Law and suffered the consequences -- Mattel's home
computer, eighteen months in the making, was priced twice
as high as it should have been. Then, as my colleagues and
I began to understand Moore's Law, we'd project the cost
curve towards $0.00 and say, "All the value's going to be
in software -- but how do you make money from software?"
It was not an easy question. Arguably, only one guy
figured it out. He realized that his big chance hung on
the fact that the first piece of software costs 100% of
R&D, while copies 2 through N are virtually free. Today
the biggest opportunity in information technology hangs on
the fact that the capacity of FTTH is virtually unlimited.
Who will be the Bill Gates, the Theodore Vail, the John D.
Rockefeller of communications networks? The opportunity is
there for the taking.
Now we return you to our regularly scheduled program of
aiding and abetting competition, deriding large clueless
entities, and open-mouthed amazement at the wonders of
today's communications technologies.
-------
Quote of Note: Dick Cheney
"There is a reason to believe that the threat level has
increased somewhat. We see more noise in the system,
more reporting that leads us to be cautious here."
United States Vice-President Richard Cheney, May 22, 2002,
in http://tinyurl.com/6ue5.
-------
Quote of Note: Tom Ridge
"I think we can anticipate more noise in the system, more
threats, because of a potential invasion."
Tom Ridge, Director of the U.S. Office of Homeland
Security, in the New York Times, March 4, 2003,
http://tinyurl.com/6u3d.
-------
Quote of Note: Paul Wolfowitz
"We have aggressive efforts underway to find new ways to
discern terrorist 'signals' from the background 'noise'
of society, but we must also recognize that enemies will
deliberately create 'noise' in the system in order to
conceal the real signals."
U.S. Deputy Secretary of Defense Paul Wolfowitz, testifying
before the U.S. Congress, September 19, 2002
http://tinyurl.com/6u42.
-------
If it's Funny, it Must be True
by Scatt Oddams
Here's one from the right -- I mean from the 'toonists down
at the CIA: http://tinyurl.com/6u13. This one's only funny
if it's *not* true.
And here's a bunch from the left -- I mean from those fine
folks who take the pledge seriously to preserve, protect
and defend the U.S. Constitution: http://tinyurl.com/4bw5.
Gotta go, Dave!
Scatt
-------
CONFERENCES ON MY CALENDAR
March 31 through April 3, 2003, San Jose CA. VON. I am
organizing a panel on April 1 (5:00 to 6:15 PM) with Tim
Horan of CIBC, Roxane "smarter-than-your-average-bear"
Googin, and Anders Comstedt, the fellow who built the
profitable, profitable, profitable, profitable, profitable,
dark fiber network in Stockholm. April 1 is one of my
favorite holidays. You will believe EVERYTHING my panel
presents -- http://www.von.com/
April 22-25, 2003, Santa Clara CA. O'Reilly Emerging
Technology Conference. My presentation will be called
Operating Models for Stupid Networks, Friday, April 25 at
2:00 PM -- http://tinyurl.com/4yhe.
June 11-14, 2002, Philadelphia PA. TedMed3. Come if you
possibly can. http://tedmed.com.
-------
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 2003 by David S. Isenberg
isen@isen.com -- http://isen.com/ -- 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
isen.com web site. The PREFERRED METHOD is to click on
http://isen.com/SMARTreqScript.html and supply the info
as indicated. The alternative method is to send a brief,
PERSONAL statement to isen@isen.com (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 isen@isen.com]
[for past SMART Letters, see
http://www.isen.com/archives/index.html]
[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. ]
*--------------------isen.com----------------------*
David S. Isenberg isen@isen.com
isen.com, inc. 888-isen-com
http://isen.com/ 203-661-4798
*--------------------isen.com----------------------*
-- The brains behind the Stupid Network --
*--------------------isen.com----------------------*