Subject: Re: 100 Acre Software...a look back
From: Jonathan Corbet <>
Date: Mon, 28 Jun 1999 16:22:14 -0600

> I don't have any particular conclusion here, except that I haven't
> seen Red Hat make a mistake yet.

Actually, there are some interesting things to note when looking at Red Hat
as a free software business.  Red Hat has done a lot of things right, they
are doing great, and I don't begrudge them that for a second.  They have
earned it.

But, if you look at their S-1 filing, one of the things that really jumps
out is that they do not see their future growth in selling boxes of free
software.  Their initiatives are elsewhere.  Web sites, enterprise support,
etc.  In some sense, they seem to think they have reached the end of the
model that has brought them this far.

Anybody who thinks they have never made a mistake might want to have a look
at what their support offerings were a year ago.  Red Hat put together a
network of "support partners" who were to be the point of contact with the
clients; Red Hat was to provide bug fix services and some others.  We
(Eklektix) were one of those partners.

The idea had its appeal.  We were supposed to be able to sell these support
contracts with the Red Hat name on them.  These contracts, while not great
revenue earners in their own right, would be the foot in the door that
allowed us to sell higher-value services to the clients.

People interested in the support side of the business might want to look at
how this program failed to meet expectations on every count.  Very few of
these contracts were sold (by any of the partners, as far as I can tell),
many of the back-end support services never materialized, and the sorts of
customers that *did* buy these contracts turned out to be extremely poor
markets for other consulting services.

Red Hat abruptly ended this program last fall.

We declined the opportunity to be a reseller of their new,
centrally-supplied "enterprise support" services, so we have no first-hand
knowledge of how it is going.  There have been indications in the press,
however, that business has been slow.  Their S-1 says that over 90% of
their revenue still comes from selling boxes of free software.

Our biggest lesson from the whole thing is that low-level OS support is not
a fun business for people who would rather be writing the Great American
Kernel Module.  As for why the program as a whole fell through, we're still
trying to figure that out.  Support was supposed to be the big thing
holding back corporate acceptance.  But the availability of that support
opened very few doors.  Maybe a year and a half ago was simply too soon?

To end this rambling thing: Red Hat has done great, and will probably
continue to do so.  But it's also true that not all they touch turns to
gold, that success in building the brand for software does not necessarily
carry over to success in other endeavors, and that they see the other
endeavors as providing their future growth.  It will be interesting to


Jonathan Corbet, Eklektix, Inc.