Subject: Re: "Reasonable Profits"
From: (Frank Hecker)
Date: Wed, 02 Sep 1998 13:14:04 -0400

Bob Weiner wrote:
> I think they and others among us are genuinely interested in
> whether someone can produce a scalable, replicable FSB model.

Indeed that is exactly what I am interested in, for the practical,
idealistic, and economic reasons I've previously stated.

I also find it a simply fascinating exercise in business strategy from
an intellectual point of view, combining as it does business, technical,
social, and political aspects in one big knotted-up ball around the
seeming paradox of making "something" (a thriving new software industry)
out of "nothing" (software without license fees).

> I have heard no discussion of exploitation or of any bias against
> FSBs, unless non-linear growth in some way bespeaks of anti-FSB
> views, which I seriously doubt.

I certainly don't intend my comments on some of these economic and
business issues as expressions on anti-FSB views on my part, in case
anyone took them as such.  I believe that considerations like liberty,
justice, and social utility lie at the very heart not only of the free
software movement considered as a community of people, but also of what
it will mean to create and sustain an industry of FSBs.  I would compare
the present time to the era that saw both the beginning of corporate
capitalism and the spread of Enlightenment views on individual freedom
and dignity and representative government; I believe both developments
were organically interdependent to a large degree, and I believe the
same is true today regarding the idea of FSBs as economic entities and
the idea of libre software and open software development as a social and
political phenomenon.

> If people are interested, I can post my notes from a talk by Darlene
> Mann, a VC, on what actually constitutes a business model.  I think
> that may open some eyes to the complexities of building sustainable
> businesses.

I for one would be very interested in seeing these.

One thing I don't think people realize at a gut level (I know I forget
this sometimes) is the scale on which the commercial software industry
operates in financial terms, and what are considered reasonable levels
of revenue or profit.  I suspect that an FSB that did $50M of revenue
per year would be considered very
successful.  However that is an order of magnitude less revenue than
Netscape at $500M+ a year, and Netscape in turn is considered a
second-tier commercial software company with an order of magnitude less
revenue than a first-tier company like Oracle.  Microsoft of course is
in a class by itself.

I don't deal with VCs, so I can't speak to what their expectations are,
but I suspect that for FSB models to be of major interest to them
they're going to have to have the potential to produce pretty fair-sized
companies in terms of revenue or profits.  (Or if for some reason FSB
models are not capable of doing that, the VCs will have to see some
other way they can get their expected return on investment.)

> I can tell you from our own two years of experience at Altrasoft in selling
> some GPLed tools to major organizations, that there is a distinct and
> reasoned bias against purchasing anything that can otherwise be obtained
> without a purchase.

This is a good point.  One advantage you have with a product with
right-to-use licensing is that you can let companies deploy evaluation
copies and then go in and help them "get legal"; with libre products you
of course do not have that option.

> Service is useful to
> corporations but in our case, selling into the technologist marketplace,
> interest is largely driven by the technology itself and if they can obtain
> that without a purchase, there is seldom enough of a perceived benefit to go
> through the purchasing cycle to get the service.

I should also note here that selling developer products is distinctly
different from selling products that touch every user in some way (like
desktop programs or products supporting enterprise-wide services like
mail) and that in turn is distinctly different from selling products
that implement a component business function (like an ERP system).  I
would say that selling developer products is hardest of all: the end
users are very demanding of quality and functionality, they are
typically only a small fraction of the total user population (so it's
more difficult to convince a customer that a true site license is
justified), and they have only limited buying authority on their own. 
As Bob Weiner notes, with libre software it's even harder, which leads
me to salute those who have been able to create successful FSBs in the
tools area.

> We have even had clients
> say that they wanted to pay only for the new features of the product that
> were not available on the net.

This sounds like the ACT situation again, and makes me wonder if some
sort of paid exclusive "first and earliest access" program could be a
justifiable feature of FSB models in general.  (This of course assumes
that you can sustainably lead the net in terms of development.)

> These are some of the issues that any FSB
> serving similar markets will face, all on top of the hundreds of other
> risk-factors that any growth-oriented business faces.  (Consumer markets are
> largely different because the user, buyer and decision-maker are often one
> and the same and there are no formal purchasing policies.)

Plus priorities are different; for example there is less emphasis on
doing continual upgrades and purchasing formal support.

For what it's worth, my gut feel is that FSB models for consumers will
end up mainly revolving around libre software used to provide access to
revenue-generating services.

Frank Hecker          Pre-sales support, Netscape government sales