SMART Letter #9:
June 23, 1998


              SMART Letter #9 - June 23, 1998
        For Friends and Enemies of the Stupid Network
            Copyright 1998 by David S. Isenberg
      This document may be redistributed provided that
      the 11 lines containing this notice accompany it. -- -- 1-888-isen-com
      It takes SMART people to design a Stupid Network


Baby Bell boss Ed Whitacre must be delighted that the
Microsoft anti-trust case is such a big media event.
RBOCs, like his SBC, have been (ahem) "exercising" near-
monopoly power since 1984. Got bandwidth?  Too bad.  Ed is
not thinking about when your xDSL service will be
installed.  He is doing what he's good at - playing
Monopoly to win. While we're taking a ride on the Reading,
SBC's buying another five states.

The bandwidth barons of new long haul networks are playing
a very different game.  Big bandwidth has arrived. A gang
of opportunistic entrants, Qwest, LCI, Level 3, and
Williams among them, see triple-digit traffic growth and
miraculous advances in Dense Wavelength Division
Multiplexing (DWDM).  They're driving exponential
expansion.  Profit margins rival illegal drugs, according
to Qwest's Joe Nacchio.  And in one scenario, the business
could be just as risky . . .

Meanwhile Ed Whitacre isn't letting technical advances
distract him from SBC's consolidation-constipation
strategy.  Megabit access means RBOC mega-trouble two
different ways:
(1)It creates a need for early capital improvements -
   Big Fast Routers and Dense Wavelength Division
   Multiplexing (DWDM) fiber - before installed
   narrowband voice equipment is fully depreciated.
(2)It lowers barriers to competition.  This year an
   NT box full of Dialogic can switch 1680 lines.  Drop a
   couple of DS-3s, or even cheaper bandwidth via OC-3,
   and poof! you're a telco.  Now IP telephony makes it
   even easier.
So from the start, Ed's motivated to keep broadband local
access scarce and expensive.

Both supply and demand are on a headlong spank.  On the
supply side, a study by KMI Corporation for the
Optoelectronics Industry Development Association estimates
that 100 times more US long distance capacity exists in
1998 than there was in 1994.  Investments are in place for
this to continue.  On the demand side, Internet traffic is
doubling between one and three times per year.  That means
that it could grow by a factor of 16 - or a factor of 4096
- in the next four years.

So if we take as given that bandwidth will grow 100 times
in the NEXT four years, and as plausible that traffic could
grow by a factor between 16 and 4096 in the same interval,
we have the ingredients for two widely divergent outcomes.
Its alternative futures time:
  (1) Bandwidth (supply) and traffic (demand) could grow
      together.  New applications and continued technology
      improvement would spur new demand. Real-time
      applications (e.g., Internet video) would appear and
      grow popular, despite relatively expensive local
      broadband access.  CATVs and CLECs would field
      broadband access via cable modem, xDSL and wireless.
      Clever schemes like caching would throttle the growth
      of demand, to some extent. Long haul bandwidth would
      remain valuable.  Companies new and old would make
      their numbers.
  (2) Bandwidth could grow faster than traffic.  New
      bandwidth from Qwest, Level 3, Williams, and LCI join
      beefed up electro-optics in AT&T, Worldcom, Sprint,
      Teleglobe/Excel and GTE networks.  Too many companies
      field too much capacity in too short a time.  Too much
      of a good thing is manageable only if a few entities
      own it.  But more players make more chaos.  Perhaps an
      especially ambitious player becomes more willing than
      others to cut prices aggressively.  Others would be
      forced to follow.  The result could be a collapse of
      long haul prices, and a bunch of weak companies
      looking for business . . . or buyers.

Each of these scenarios has their adherents in the press.
Andy Kupfer writes, "Network bandwidth will always be a
scarce commodity," in Fortune (7/6/98, p. 92 ff).  Bob
Metcalfe wonders if "the Internet's traffic bubble will
burst," leaving all that bandwith with nothing to carry, in
Infoworld (June 15, 1998, p. 167).  Me, I'm agnostic, but
healthily skeptical of any trend that attracts multiple
powerful players before any revenue develops.

Abundant, reasonably priced local broadband access by the
incumbent local carrier would throw the balance towards the
first alternative outcome.  Available affordable access
could drive new, nifty broadband applications.  (Imagine
how inexpensive, popular, high quality, broadband video -
live, on demand, volunteer, candid, or what have you -
could change the Internet.)  This would make more end
users, more long haul traffic, and a vastly more compelling
and useful Internet.

But a more powerful Internet would weaken Ed Whitacre's
hand. No wonder he's not interested in your xDSL. If local
access is scarce, long haul traffic will be too.  He's
rooting for outcome #2, for flailing start-ups, for a price
collapse. He'd rather be a buyer of weak long haulers than
bait for the bandwidth barons.

I'd be amazed if any of the Baby Bell boses - Whitacre,
Smith, Notebaert, Ackerman, or Trujillo - steps up to local
broadband service in any meaningful way.  It's up to the
entrepreneurs.  As Bill Gates put a billion into Comcast
data upgrades, let another *hundred* billion flow into HFC
and cable modems, into xDSL by savvy CLECs, into spread
spectrum, and satellite, and LMDS, and powerline carrier,
and fiber to the home. (OK Roland?)  The incumbent
Interexchange Carriers might come to realize that this is
their surest passport to the future, surer even than the
value of their own backbones.  Content providers might
benefit from stepping away from yesterday's business model,
even if only experimentally, to realize that the future
value of their content depends on it. Equipment providers
will sell more equipment on the edges and in the core too.
And surely the new bandwidth barons will appreciate that
their daring new investment is vastly more valuable in
Scenario 1.

There is a huge constituency for a strong, vital,
broadband, continuously evolving Internet.  In fact, I
can't imagine anybody with vested interests on the Scenario
2 side, except for RBOC employees, state PUC officials with
kids in college, Senators and Congresspeople, and
stockholders.  To the latter, sell.

(Whether or not we believe the scenarios above, and there
are many reasons why we shouldn't, SMART people are invited
to post questions, flames, assertions and doubts to - and read others' input on

SMART Person Brian Mulvaney is a product
manager for a company that sells tools and services to
companies working on Y2K.  I find his prose spare and clean
and his viewpoint rare and keen.  He granted permission to
publish the following excerpt here:

      ". . . . The number of actual date defects they've
      found has been surprisingly small--though quite a few
      of these have been subtle, insidious bugs that could
      have wreaked havoc with the bank's data.

      "The overall effort has been a hard slog and no fun at
      all.  Many IT staff members will probably leave
      afterwards, some have already left.  And at the end of
      this enormous effort the bank will be left with
      systems that are essentially unchanged with the sole
      redeeming benefit that they should (knock on wood) be
      millennium compliant.  Return on Investment = Zero
      All that money and all those man years just to
      maintain status quo.  I'm suspect that their IT
      Director (CIO) has alternating nightmares:  one night
      it will be that his negligent banking competitors that
      barely spent on the problem will sail into the
      millennium essentially unscathed; the next night it
      will be that despite having compliant systems his bank
      will be dragged under by the failures of other banks
      and big borrowers.  Either way the end result would be
      a valiant effort that ended in futility.

      " . . . Here's an outfit that started early enough,
      treated the problem as a business issue rather than a
      technical issue, did most of the right things, made
      sacrifices, behaved competently and responsibly and
      spent a fortune in time and money.  Yet they still
      aren't totally out of the woods and don't know what,
      if any, payback there will be.  Is it any wonder that
      many (most?) organizations don't appear to be very
      serious about addressing Year2000?

      "I keep trying to find a silver lining somewhere.  My
      best hope is the Augean Stables outcome:  The
      Millennium might flush out a monstrous load of crappy
      information systems and outdated business practices
      that are bound to them.  Assuming we make it through
      the experience with the essentials intact, we might
      actually be positioned for a stunning economic boom.
      Seems like there should be a hook into dumb networks
      here, somewhere.  Basically, dumb business practices
      will prevail over centralized, intelligent ones
      because they are less bound to outdated IT
      infrastructures, that dumb networks are the conduit
      for dumb business practices."

Thanks, Brian.

By my calculations (which have been known to slip a decimal
point or ignore an obvious factor) a single OC-192 can
carry 150,000 phone calls.  There are 150 million phones in
the US.  Say 10% of them are off hook during "busy hour."
One single glass fiber, using 100 times OC-192 DWDM, could
carry all of these calls, local and long distance -- *IF*
they all shared the same route for an instant.  Is
something wrong with this picture?

Sven-Christer Nilsson, Ericsson's CEO, knows his customers.
A couple of weeks ago at Jeff Pulver's VON Europe, he said
that the Internet needed to have reduced latency, improved
reliability, security and billing "to be useful." Certainly
he MUST have meant "Useful to Ericsson's customers."  The
Internet is probably useful to somebody, Sven-Christer.
After all, it has been growing at 200% to 800% per year for
some 20 years.  But this has been largely without the
participation of Ericsson's customers, the world's main
line telcos.

In-region Long Distance entry used to be the 1996 Telecom
Act's carrot to encourage opening the local loop for
competition.  But so far no Bell has won an open loop
prize.  Now the motive is gone - a "sit on your hands"
clause in the Act lets the Babies do in-area Long Distance
after 36 months - that'll be February, 1999 - even if their
monopoly is intact.


Washington DC, July 9 & 10, 1998:  Next Generation IP
Networks, a Jerry Lucas Telestrategies Meeting.  I will be
speaking on July 9, right after Lee McKnight, editor of the
mighty tome "Internet Economics," who will be a hard act to
follow, I am sure.

Chicago IL, July 27-29, 1998:  Profitable Applications for
Intelligent Networks, an Institute for International
Research event. I will "debate" John Young, an IN
proponent.  Hmmm. or call Meredith Gendell
on 212-661-3500 x3113.

New York City, September 1-3, 1998:  The Mongo Mondo
Telecom Business Show at the Javits Convention Center.  By
my friend, entrepreneur Marc Ostrofsky and his merrye band.
I'm looking forward to hearing Dave McCourt, RCN's boss and
Steve Rosenbush, USA Today's constantly surprising telecom
reporter.  I'll be talking at 4PM on Sept 2.

Lake Tahoe NV, September 15-17, 1998:  DON'T MISS George
Gilder's TELECOSM.  Guaranteed new perspectives at this
one, folks!  You've gotta hand it to Gilder, he's
completely immune to CW, walks 100% of his talk, and more
than occasionally he is mind-blowingly right on.

markets are open to scores of competitors today, thanks to
the aggressive efforts by Bell Atlantic and other regional
companies.  Meanwhile, the nation's long-distance markets
are held in the iron grip of AT&T, MCI, and Sprint."  A
Bell Atlantic press release quoted in Forrester Research's
December 1997 Telecom Strategies Report.  Thanks to SMART
Person Chris Mines.
Be SMART at the periphery,
David I
<<to unsubscribe to the SMART List, send a brief
unsubscribe message to>>

<<for past SMART Letters, see>>

David S. Isenberg     
18 South Wickom Drive   888-isen-com (anytime)
Westfield NJ 07090 USA  908-875-0772 (direct line)
                                908-654-0772 (home)
     -- Technology Analysis and Strategy --
        Rethinking the value of networks
      in an era of abundant infrastructure.

Date last modified: 22 July 1998