Don Gilbert gilbertd at sunflower.bio.indiana.edu
Thu Jul 28 11:41:20 EST 1994


I believe you are wrong about a few thousand US$ being an excessive price
for software developed for a niche market like molecular biology.
Widely used commercial packages like wordprocessors can afford to sell 
for few hundred dollars because of the hundreds of thousands or millions of
packages sold per year.   The expenses to produce a good piece of software
and especially the ongoing expenses to market it (e.g. to make sure some of
you actually buy it) are very high, regardless of whether you sell 100
or 100,000 copies.  When I was an independent software developer and
tried selling my software commercially, the advertising costs alone
were enough to keep me from getting anywere: several thousand for one
small, back of the magazine ad (the kind almost no-one reads) in one
issue of one computer magazine.

The molecular biology software companies do not sell 10,000/yr copies
of their software.  The market isn't that large.   I don't really
know the size, but I'd say 2,000/yr was a high number for a popular
molbio program.  Let us say it sells for US$3,000.  That yeilds US$6,000,000 
gross income.   Of that normally more than 50% will go toward
advertising and marketing (required to generate that many sales!).  It 
may well be that 70% to 80% of gross income goes to marketing expenses
to generate those sales in the software business.  Have you ever
looked at the profit margins on some of the big software companies?
If I recall, Borland had sales around $150 million a few years back with
a profit of only a few hundred thousand (and they may well be losing
money this year).
Any complex software today requires many programmer-*years* to develop.
Estimate $100,000/developer/yr including fringe benefits (this is probably 
a low estimate, but more than I earn:),  with an estimate of 6-person-years 
to develop (widely used commercial packages would involve at least 10s 
of person-years if not 100s).  Add in the cost of various technical 
writers, support personnel, etc.  It adds up to where any profit is only 
a small percentage of the total income.  And if a given package isn't 
popular, you loose money on it (paid for from profits on a popular 
package).  Only if you sell lots more than expected do you see 
significant profits.

This is all my rough estimation from the time when I was a very poor
independent developer.  Some software companies do make profits.  But
the mol-bio market is such a small one, and the required price for
software packages to meet the basic marketting and development costs
are so high, that it will remain a risky and uncertain market in my
opinion.  I don't have hopes of selling any of my software in this

-- Don

-- d.gilbert--biocomputing--indiana u--bloomington--gilbertd at bio.indiana.edu

More information about the Bio-soft mailing list

Send comments to us at biosci-help [At] net.bio.net