Review of Charles Ferguson, High Stakes,
No Prisoners
J. Bradford DeLong and A. Michael Froomkin
October 1999
DeLong is a professor of economics at the University
of California at Berkeley, and a research associate of the National
Bureau of Economic Research.
Froomkin is a professor of law at the University
of Miami Law School, and a non-executive director of the Internet
startup company Out2.com.
A link to Charles Ferguson (1999), High
Stakes, No Prisoners (New York: Times Books: 0812931432)
at amazon.com.
This is an early draft of a book review that will bve forthcoming
in the Harvard Business Review.
Charles Ferguson has written a very honest book. That honesty
is one chief reason to read it: he dishes the dirt on Netscape,
Microsoft, his lawyers, his venture capitalists, and not least
himself. But his very honesty gives the reader some critical
distance--and gave us the tools to question how long the core
conclusions of the book will continue to apply.
In 1993 Charles Ferguson--MIT-trained engineer, consultant,
and high-tech industry analyst--had a brilliant idea: the world
needed a visual development software tool to create online information
systems. The tool had to be visually-oriented to be useful to
the non-programmers who knew the information. Yet the tool had
to be sophisticated to allow organizations to structure their
data in useful ways. Ferguson sunk his then-life savings into
his idea. He created his software corporation, Vermeer. With
his partner, Randy Forgaard, he assembled a very good programming
staff. He raised venture capital. He pursued the enterprise with
monomania mixed with paranoia. And by the end of 1995 there was
code that was more sophisticated than the code of potential competing
programs like NaviPress, Netscape Composer, or PageMill, and
that actually ran.
Ferguson and his startup jumpstarted an important part of
how we interact with the Internet. Because of his idea, because
he backed it to the hilt, and because others saw what he was
doing and imitated him, we as a society got a nine-month head
start on the development and diffusion of visual tools for Internet
information systems design. This has enriched us all: we are
in his debt. And he managed to carve off for himself a small
slice of the user surplus that he and his organization created,
and become rich.
At the end of 1995 Microsoft offered somewhat more than $100
million for Vermeer, for Vermeer's key technical personnel--and
for Ferguson's silence for two years. Ferguson and the rest of
Vermeer's board accepted. His software project was renamed "Microsoft
Frontpage."
The two years of silence have elapsed. This book is Ferguson's
story of the Internet startup that made him rich. Or, rather,
the story of the Internet startup Vermeer is one of three major
threads that make up this book. For there are really three essays
of varying length woven together within one set of covers:
The first essay is the story of Vermeer and of its founding entrepreneur,
Charles Ferguson.
The second essay is the story of the rise of the Internet and
a briefing on the dilemmas of high-technology corporate strategy,
written by engineer-consultant-business analyst, Charles Ferguson.
The third essay is an attack on the harmful effects of Microsoft,
written by someone--Charles Ferguson--who feels guilt at having
profited from and added to Microsoft's economic leverage.
Other topics ranging from the Microsoft anti-trust trial to
privacy law to
productivity theory are raised, briefly discussed, then dropped.
One of the more interesting asides is Ferguson's analysis of
the sources of today's high technology: although the development
came from the private sector, in his view the research came and
had to come from the government. One of the more substantial
tangents is a ten-page attack--the particulars of which seemed
to go awry in places (or so says DeLong, whose office is on the
same hall as two of Ferguson's targets)--on the Regional Bell
Operating Companies and the economists who consult for them.
This is a book rich in asides.
It is also a book that is less than completely successful.
The essays braided together to make up the body of this book
are not quite the best in their respective classes.
First, for entertaining and informative entrepreneur's eye
view stories about business enterprise in the computer age, Andrew
Grove's _Only the Paranoid Survive_ gives a more nuanced and
deeper view of the industry. The pseudonymous "Robert X.
Cringely"'s _Accidental Empires_ does a better job of placing
the developments that made today's high-tech what it is in their
full and proper perspective.
Second, if you want a briefing on the dilemmas of high-technology
corporate strategy, Carl Shapiro's and Hal Varian's _Information
Rules: Competitive Strategy in the Information Age_ provides
a better analysis of what high-tech industries are like, and
what are the features that make competition in these industries
different and interesting. To a significant degree this is simply
a problem of space: Ferguson is cramming three books into the
space of one. To some degree this is a problem of narration:
Ferguson sees the world in sharp black and sharp white only.
There are no shades of grey, no tolerance of ambiguity, no recognition
that things look different from different points of view. To
his great credit Ferguson recognizes--as few analysts are ever
adult enough to do--that he has been very certain yet wrong in
the past. Yet this recognition of error in the past does not
lead to a recognition of uncertainty and ambiguity in the present:
people are geniuses or idiots, corporate strategies are brilliant
or misguided, organizations are healthy or fatally diseased.
Third, the Microsoft essay is too short to be fully convincing.
You can get Ferguson's point of view much more depth and thicker
description by reading _Hard Drive_ or _Competing on Internet
Time_. Moreover, Ferguson's Microsoft essay is somewhat limited
by the fact that he adopts a point of view too close to that
of Microsoft's competitors and not close enough to the broader
view of consumer surplus and the public welfare. Ferguson feels
sorry for Borland and Netscape because Microsoft gave them "cashectomies":
giving away the equivalent of their flagship products for nearly
free (in the case of Borland) and completely free (in the case
of Netscape). But from a public interest standpoint it is hard
(though possible) to complain vociferously about the result:
the bird in the hand is that lots of users got lots of good software
very cheap, while the value of those still in the bush is unclear.
As we argue below, there is at least a case to be made that Microsoft
may not retain long-run market power through its software giveaways.
So if the three main threads are not quite best-of-breed,
why read this book?
People should read this book because Charles Ferguson is an
entrepreneur *and* a business consultant *and* a policy analyst.
He thus has a very interesting multiplicity of vision. Ferguson-the-entrepreneur
launching the Internet startup Vermeer is watched in the back
of his head by Ferguson-the-industry-analyst, who is thereby
led to rethink his views on the relative comparative advantage
of large U.S. companies vs. the Silicon Valley system. Ferguson-the-entrepreneur's
ideas about what Vermeer should do are powerfully and explicitly
influenced by Ferguson-the-industry-analyst's views on how to
create an entrenched market position in the information age.
Usually industry analysts and consultants have little clue
about how difficult it actually is to run a business. Usually
entrepreneurs have a hard time expressing the vision of industry
dynamics that they intuit. They cannot--even when successful--easily
explain just what it was that led them to follow the strategies
for creating market position that they adopt. Ferguson can, and
does. Usually policy analysts have little grasp of how the private
sector really works, and of what government interventions in
the market will actually do. Ferguson-the-entrepreneur has a
solid grasp of these issues.
The combination gives the book a combination of a wealth of
perspective with a form of street credibility that few books
about the Internet economy possess, and makes it worth reading.
Thus, for example, when Ferguson tells us that government R&D
people are "light years ahead" [59] of their counterparts
in industry in both brainpower and vision, that conclusion cannot
be dismissed as the rambling of some academic consultant who
has never run a business or taken a product to market.
And there is all the dirt Ferguson dishes. It is very entertaining
to watch it fly.
Vermeer
In early 1994 MIT-trained engineer, management consultant,
and computer industry-watcher Charles Ferguson had the spectacularly
good idea that extraordinary utility could be created by the
combination of:
--a common information server architecture.
--a common information browser architecture.
--a visual development tool that would allow organizations
without programmers to structure the data they wanted to place
on their information server to make it accessible and useful.
As Ferguson quickly discovered, the first two parts of this
combination, the http protocol and the web browser, had already
been created by Tim Berners-Lee of CERN, the inventor of the
world-wide-web. The fact that they were open-source--that their
specifications were published, that any programmer anywhere in
the world was free to try to improve the running code, and that
any programmer anywhere in the world was free to try to interface
with or extend the system--was key to the explosion of the world-wide-web.
By going the open-source route Berners-Lee traded away a small
chance of becoming a billionaire and in return received a role
as the human whose work has had the greatest consequences for
the world of anyone in this decade. And the open-source success
of the web made it impossible for any alternative, proprietary
wide-area information architecture to succeed. As Microsoft CEO
Bill Gates recognized early in 1995, the open-source nature of
the web doomed Microsoft's launch of the proprietary Microsoft
Network [MSN]: as he wrote, Microsoft had no answer to information
distributors and users who asked "why they should use MSN
instead of just setting up their own web server."
But Ferguson and his partner, Randy Forgaard, recognized that
the third part was absent. The wildfire spread of the open-source
web greatly enhanced the potential value of creating the third
part--a powerful visual development tool for building and maintaining
world wide web sites. By late 1995 it was clear that the Vermeer
team had mostly succeeded. In large part because of Ferguson's
monomania, they had a shipping product, a sophisticated visual
web development tool, six to twelve months ahead in functionality
of all of its major competitors. They had an excellent good programming
staff with experience at this slice of business. But by this
time Ferguson loathed his CEO and his VC backers.
Moreover, Vermeer did not have a functional marketing organization.
Ferguson reports that during the three months between Vermeer's
launch of its product, FrontPage, and its acquisition by Microsoft,
it sold less than 300 copies--a rate of total sales of less than
$600,000 a year for a company with a then-burn rate of perhaps
$10 million a year. Thus by the end of 1995 the only live option
was to sell out to a software company that did have functional
sales and distribution organizations.
Of the companies that might have acquired Vermeer, AOL had
already bought a company in Vermeer's line of business--Navisoft,
with the NaviPress web development tool and with NaviServer (now
called AOLserver, a now open-source product that many consider
to be, along with open-source Apache, the best-of-breed in webserver
programs). Adobe had bought PageMill. Netscape and Microsoft,
however, were still thinking about acquiring web development
tools.
Entrepreneur-Ferguson sold Vermeer to Microsoft for $130 million,
in spite of qualms high-technology policy analyst Ferguson felt
about the social utility of adding a leading position in another
market segment to Microsoft's assets. What did Microsoft get
for its $130 million? It got a nine-month head start in web development
tools embodied in the code and in the technical team that it
acquired.
The story of Vermeer is interesting and entertaining. However,
the lessons learned will probably not be of much use to others
in similar situations who are now trying to obtain venture funding
and rapidly expand their organizations. The book contains a lot
of true knowledge about how one bargains with venture capitalists
for funding. But this is one case in which information is not
(much) power: the VCs have the money, and you need it. Ferguson's
lament that VCs do not compete against each other because they
know they will deal with each other again may be tempered slightly
by the growth in the number of funds seeking to bankroll startups,
but the basic power imbalance remains so long as the number of
would-be startups grows at least as quickly.
Software Markets
As Vermeer walked its path toward launching its product, Ferguson
focused on two things. First, he focused on trying to get the
programmers to design and create the best possible visual development
tool for creating and maintaining websites. Second, he focused
even more on trying to design a visual development tool that
would make money for the company.
The second is an extremely difficult task because of the features
of software markets.
The first is that in general in any software niche there can
only be one highly profitable product: the one with the largest
market share. Costs of manufacturing software are almost all
up-front, one time, costs. The production economies of scale
are massive because software is what economists call non-rival.
One master copy on a disk can be duplicated essentially for free
and run by any number of users. These supply-side economies of
scale are matched by demand-side economies of scale as well.
If your co-authors, your supervisors, or your subordinates use
Microsoft Word, you have to use Microsoft Word as well unless
you want to waste half your life dealing with file-format-translation
glitches.
Given the relatively small size of the consolation prize,
it is no wonder that the drive to become *the* standard becomes
all-consuming. Software makers like Vermeer become obsessed with
speed-to-market, secrecy (to prevent competitors from recognizing
what their product will be), and initial market share. The need
for dominant market share also encourages software bloat. As
Ferguson perceived it, "You can't become the industry standard
unless your product covers every major portion of the market."
Success in this strategy acts as a barrier to entry. Given
the high fixed costs of making a compatible and multi-functioned
product, competitors may often will be discouraged from even
entering the market segment if there is an entrenched early mover
with a large market share. Indeed, even a firm willing to commit
substantial resources may find that a pre-existing proprietary
standard cannot easily be dislodged. Microsoft, after all, found
it expedient to acquire FrontPage rather than assign a team of
programmers to reinvent it. An open source standard is even harder
to dislodge or control: Bill Gates originally hoped that Microsoft
could make its Word document format displace the HTML format
as the standard for web-based publishing, but soon had to admit
that HTML was not going to go away.
Thus Vermeer and its software designers had a big problem.
They had to create a version 1.0 that was good enough to rapidly
achieve a dominant market share. They also had to have plans
for versions 2.0 and 3.0 that would make users wish to pay to
upgrade. Without upgrades and revenue from upgrades you do not
have a future revenue stream. And they had to--somehow--lock
consumers into FrontPage, so that when they did wish upgrade
they would find it very expensive to upgrade to anything other
than the latest version of FrontPage. (Ferguson is acid about
Netscape's failure to appreciate the need to lock-in--or even
identify--its customers, and attributes much of their difficulties
to this.)
Microsoft and Open Source
Perhaps Ferguson's most successful long-run contribution from
the standpoint of Microsoft's bottom line was to manage to create
a software architecture that did achieve a substantial amount
of lock-in. The server-side extensions of Microsoft FrontPage
do indeed make it very expensive for users who have utilized
these advanced features to shift to any other program. Moreover,
it appears to be difficult to install the server-side features
on any machine that is not running Microsoft's Internet Information
Server--there appears to be lock-in not only over time, but between
different Microsoft programs as well.
But this lock-in makes users' lives more difficult: it is
definitely *not* in users' interest for such lock-in to be successfully
created. It is, Ferguson thinks, definitely not in users' or
in America's interest for Microsoft to take its two current effective
monopolies--in operating systems and office productivity suites--and
add to them three more monopolies, in browser software, in web
development tools, and in server software. Yet when Ferguson
looks back on the impact of Vermeer, he sees it as having accomplished
three things: (a) made him and his people rich, (b) jumpstarted
the web development tools industry, and (c) given Microsoft a
very good chance of extending its dominance into new markets.
And this third worries policy-analyst Ferguson very much, for
he thinks that the world is ill-served by a Microsoft dominant
across many different software markets.
Ferguson sees Microsoft's business model as based on two principles:
(i) establish and control its own industry standards, and (ii)
commoditize every other related business. The first is accomplished
through embrace-and-extend. Microsoft initially embraces a standard
proposed by others. Then it cuts its price to or beyond the bone
in order to gain a significant market position. Then it extends
the standard in proprietary ways that its competitors cannot
easily match. And it winds up with an effective monopoly. The
second is accomplished by Microsoft--in cooperation with Intel,
and usually some PC manufacturers--making sure that new hardware
functions are standardized in an open, non-proprietary way. This
produces a hardware industry with a high degree of interoperability
that is ruthlessly competitive. And because competition means
that users spend less on hardware, they have more to spend on
Microsoft.
Hence his policy recommendation: dismember Microsoft in a
way that prohibits it ever getting back together, or any piece
of it ever recovering its present market position. If this is
not accomplished, Ferguson fears, progress and innovation will
slow as Microsoft ossifies, and we all will be the poorer for
it.
Ferguson's fear of Microsoft is built on the insight that
"the key prize in high technology is proprietary control
of an industry standard" [24], an insight that he arrived
at in 1992. It was this insight that led him to place so much
stress on protecting intellectual property and creating lock-in
while he was running Vermeer. It worked, and made him rich. Call
this the "Microsoft Vision," for it was the proprietary-standard
lock-in possibility created by Vermeer's server side extensions
that made Vermeer worth so much to Microsoft. (Moreover, there
was the possibility that perhaps the server side extensions could
be made to work more easily and effectively with Microsoft's
Internet Information Server and other Microsoft software than
with competing products.) Thus Ferguson's current fear of Microsoft
is the product of much the same mindset that made Microsoft think
that Ferguson's company was a valuable acquisition, for Microsoft
certainly believes in the Microsoft Vision.
Yet perhaps this insight has outlived its usefulness. Perhaps
what worked--gloriously, for Ferguson--in the 1990s will not
work ten years from now. In reading _High Stakes, No Prisoners_
we were frequently struck by a strange silence. As we at least
look at Microsoft today, other than the possibility of action
by the Justice Department, its principal competitive threat and
restraint--in the operating system business at least--is not
Apple or Sun but is the open source Movement and its freeware
operating system, Linux. That open source vision may or may not
come to provide serious long-run competition for the Microsoft
Vision. It may or may not provide long-run constraints on Microsoft's
profitability and market power. But it seems fair to ask that
any discussion of market structure and competition policy in
the software industry address the issue of competition from Open
Source software.
All this is not on Ferguson's radar screen at all.
The Open Source movement has interesting properties that suggest
to us it is worth taking seriously. Essentially volunteer software
development would seem particularly vulnerable to the tragedy
of the commons, but open source has evolved a number of strategies
that at least ameliorate and may even overcome this problem.
One is the now-familiar ways in which open source authors gain
status and participate in a sort of online gift-exchange. Perhaps
less well understood, but ironically well-revealed by _High Stakes,
No Prisoners_ itself, is the way in which participation in an
Open Source development process provides a strategy for undercapitalized
or risk-averse software designers who do not themselves wish
to be in the open source business. Using an open source process
may limit your upside but it also protects your flank. As Microsoft's
failed attempt to replace html with Word's .doc standard demonstrates,
once an open source standard takes root even a well-financed
proponent of a propriety solution may be stymied.
Similarly, from the user perspective, open source provides
at worst a mixed blessing. The absence of the prospect of an
enormous payout may retard the development of new features, and
also reduce the richness of the feature set. However, since most
people apparently use a fairly small subset of the features provided
by major packages open source may make proprietary designer add-ons
both technically feasible and economically rational. Open source
may also revise slower, arguably a blessing these days, but it
tends to be very stable (e.g. Debian). Customer support may be
no worse under open source, as provided by a combination of dedicated,
even obsessed, volunteers (sometimes the designers themselves)
and value-added aftermarket commercial opportunists.
Indeed, one of the great strengths of the open source movements,
albeit one not always endorsed by its hard-core devotees, is
the extent to which it creates opportunities for market-driven
complementary products and services that are much more difficult
to provide for proprietary solutions. Thus, for example, Red
Hat can make a successful business providing easy to install
Unix and technical support in a way that is all but inconceivable
for a supplier of Windows 98 assistance.
Ultimately, Open Source stands a good chance to provide a
counterpoint to the Microsoft Vision because it offers a different
and apparently viable solution to the "problems" of
high fixed costs, of non-transparency, and even "non-excludibility".
The fixed cost problem is solved by spreading the cost to all
the participants in Open Source development. Since no one can
control the standard, and the point of the exercise is to allow
competing products to bloom, there is no need for secrecy, and
the paranoia level can go down. All providers in the software
market are assured that their access to it will remain. The non-transparency
problem is not eliminated, since predicting the future is as
hard as ever, but users can be reassured that the openness of
the source means that if they desperately need a feature they
can pay someone to code it for them.
Open source even alleviates what economists call the non-excludiblity
problem. In normal software markets the manufacturer has a very
difficult time making sure that only those with valid licenses
are running programs. Sophisticated copy protection systems go
haywire enough and cause enough hassles for legitimate users
that they have all but vanished in the course of the past two
decades. And in the absence of such copy protection schemes manufacturers
are reduced to relying on users' respect for manufacturers' intellectual
property rights to produce a consistent revenue stream. In open
source, the revenues cannot come from traditional intellectual
property rights alone. One must provide some sort of other value,
be it help desk, easy installation, or extra features in order
to be paid. The danger that one may not be paid at all is the
Achilles heel of the Open Source vision, and the reason why we
are not prepared to say it will necessarily predominate.
Ferguson's blind spot about Open Source software is surprising
since his entire business--his entire business model--depended
on Open Source software several times over. As he notes, critical
to the development of the Internet was the adoption of servers
based on the UNIX operating system [53]. The world wide web as
we know it exists because Tim Berners-Lee Open Sourced html and
http. (It may be telling that Ferguson found Berners-Lee "unrealistic"
when he first met him, and slams the W3C consortium that Berners-Lee
joined as a "rather useless, nonprofit Web standards group.")
Yet as Ferguson also recognizes--were there a Nobel Prize for
computer science, he would award it to Berners-Lee--the market
for his web design tool existed only because Mosaic and Netscape
web browsers, and Httpd and Apache servers, were being given
away for free.
Conclusion
There is a sense in which the characteristics that made Ferguson
a successful entrepreneur both strengthen and weaken this book.
They strengthen the book by providing street credibility and
consistency checking. Ferguson-the-business-analyst gives Ferguson-the-entrepreneur
a much more intelligible voice than enterpreneurs possess by
providing both a framework to organize the narrative and clear
well-written prose. They weaken the book because they lead Ferguson
to be so d--- sure of everything, to see sharp lines between
black and white. An intolerance for ambiguity, a refusal to recognize
uncertainty both beforehand and in retrospect--these are characteristics
that made Ferguson an excellent and decisive entrepreneur. But
they also make readers of the book unsure as to just how strongly
they should hold the conclusions Ferguson reaches, and unsure
just how strong the evidence for those conclusions is. Does he
really want us to believe that in 1995 Janet Reno had no one
working for her who understood high-tech and Microsoft?
Ferguson reports that his track record as a consultant and
policy analyst has been uneven. He thought at the time that IBM's
hire of Lou Gerstner was a disaster, he greatly underestimated
the value of the Silicon Valley system, and he overplayed the
Japanese threat to America's comparative advantage in high-tech
industries. Will his next book, in a decade or so, begin with
an admission that he failed to recognize that Open Source had
achieved a state of maturity that vastly undermined the value
of "proprietary control of industry standards"?
We are not sure. But we find our confidence in his analytical
judgment is somewhat reduced by the fact that he doesn't find
Open Source worth thinking about. And we also find ourselves
somewhat disappointed: for we are very curious to learn what
he would have to say.
Unpublished Book Reviews
|