[PLUG] Linux support business

Keith Lofstrom keithl at kl-ic.com
Thu Oct 2 11:25:02 UTC 2003


This is an elaboration of an earlier suggestion about forming a Linux
support company.  As others have pointed out, places like Linuxcare
haven't managed to fill this niche, and places like Stream are hell
to work for.  The first question to ask before starting any unique
venture is "why hasn't this already been done?"  If the answers seem
to involve dark conspiracies or mass evil, then this indicates a need
for psychiatric help, not a business startup loan.  I won't go into
details of starting a business, there are books out there for that,
and if we can't roughly agree on that information it is a waste of
time to proceed here.  So let's proceed on the assumption that normal
business principles ( legality, profitability, salability, productivity,
scalability, tolerability, etc. ) apply.  If you want to argue with
those, go start your own business, succeed, and write your own books.
At least, start your own thread about the wickedness of commercialism;
we are focused on a different problem here.

Since this is an innovation in software, I would suggest reading the
books "The Inmates are Running the Asylum" by Alan Cooper, and "The
Innovators Dilemma" by Clayton Christensen. The first book is about
product design for use, with a focus on software;  the second is about
how "cheap plus good enough" takes over existing niches.  You can find
out a lot more about these two books by Googling. 

A business solves the problems that others perceive, in return for
their money.  While that process can get mighty distorted, you do
not succeed by betting against it.  So the question is, whose money
do we want, and what problems do they have that we can solve?

Companies and rich people have money.  We want them to give us some
of that money.  They also have problems, some of which have Linux
solutions.  We can solve those problems.  At some point, we can also
solve consumer problems, but consumers have more varied problems and
their individual piles of money are smaller and more trouble to get
at.  It will take a fairly large service company to serve consumers,
and we aren't there yet.  However, when we go after the big piles
we face some competition;  we have to look for those piles where we
face less competition.  If we can manage to keep those simple facts
in mind, we can build a business.

I suspect the biggest available piles with low competition are 5-50
employee companies in the category "other".  They are big enough to
have a company network, there are no targeted software solutions
(restaurants, medical offices, auto shops, etc. already have solutions,
though weak ones), and chances are the owner or manager also spends
time taking care of the damned computers - when they should be growing
the business, so they have more money to pay us.  

So the "avatar" customer (see Cooper) is a middle-aged small entrepreneur,
fairly tight with money, who wants to get more out of her computers
with less time spent.  She doesn't want problems, but she doesn't care
about manuals and RFCs and standards (that's OUR problem).  The company
must be able to share data with suppliers and customers using Windoze,
Mac, etc.  The employees use email and the web (browsing and website),
and most are not computer-hygenic.  While she would probably like
specialized software focused on her individual business, it isn't
there yet.  She probably needs Quickbooks running under Wine, so we
can't afford ideological purity.

All that said, we can keep the machines up, provide basic applications, 
keep viruses out, write little functional scripts to solve point tasks,
and do 24x7 maintenance far more effectively and cheaply than she can
get the same things done with a Windoze solution.  Maybe we don't have
the chrome, or the games, or the industry-specific apps (though many
of those can run under Wine or Win4Lin or even offsite).  We cannot
give a flashy demo, so we will have to trade on our strengths -
FREEDOM, availability, adaptability, and - potentially - attitude. 

Attitude will be the big challenge - software people are intellectually
smart, and this leads to bad attitudes and ego.  Nobody pays for that.
In years past, Microsoft displayed slightly less attitude during the
critical growth period of the industry.  Because they were slightly
less disgusting for users than their competitors (or at least convinced
people of that, which amounts to the same thing in business), they
captured the desktop, and captured the interchange standards, and now
own the commons.  So they are now back to the "screw you, luser"
mentality, and act like they are invulnerable.  This is a hell of an
opportunity for folks who are smart enough to write and support code
that turns lusers into winners.  I think, with some attitude adjustments,
we are those smart people.

The first thing needed are enough friendly, generalist problem solvers
to form a 24x7 2x redundant first-response phone/web team.  Asterisk
might make a good rallying point.  Can Squeak Smalltalk be made to run
as a background process, and drive browser windows?  I understand Smalltalk
is good for rapid prototyping, which will be needed for numerous point
solutions and quick-fixes.  The day team will need folks that can meet
with customers, learn their needs, and code them rapidly.  Of course,
we will have VPN tunnels into customer machines, and some fairly
fine-grained permission systems to maintain customer security.  That will
need some backroom geeks, but not many.  As things evolve and the process
scales, we will need to develop securely accessable customer profiles, so
strangers can help strangers.  And we will need ways to profitably exchange 
point apps and code modules with other service providers, so we can
maintain profitability as we expand downwards towards the complexity
of individual consumers. 

Well, that is an awfully long ramble.  Perhaps we can discuss it more
tonight, at the Lab.

Keith
    
-- 
Keith Lofstrom           keithl at ieee.org         Voice (503)-520-1993
KLIC --- Keith Lofstrom Integrated Circuits --- "Your Ideas in Silicon"
Design Contracting in Bipolar and CMOS - Analog, Digital, and Scan ICs




More information about the PLUG mailing list