Time Expired Software!

Andrew Dalke dalke at ks.uiuc.edu
Mon Jan 27 18:52:33 EST 1997


In article <32EC39F3.46D2 at gab.unt.edu>, "C. J. Fields" <cfields at gab.unt.edu>
writes:
|> Susan Jane Hogarth wrote:
|> 
|> > > Justify the price for PC-GENE, which, to be quite frank, is a piece of
|> > > crap (in my humble opinion).
|> > 
|> > Easy - people _buy_ it!
|> > 
|> > sheesh, when do we get to the _hard_ questions? ;-)
|> 
|> Sorry, that doesn't justify the price; that just means some people are
|> suckered into buying it.  I probably won't be convinced unless one of
|> those commercial sales reps gives me an itemized list.  Anyway, nice try
|> but no dice.
|> 


Its a free market.  The only thing that _can_ justify the price is that
people will buy it.  Features (your "itemized list") just make people
more likely to buy it.  I vote full dice for the answer.


> I have as of yet to get a real response on why these programs are so
> high priced?  Anybody else got a response? ;->

  Okay.  Suppose I am company Mol<whatever>.  Based on our work here,
I need about $100K/ year to support a good programmer.  (Figure on
$70K/salary and benefits, $20K/yr for hardware and support, and $10K/yr
for misc.  Maybe two programmers to get $170K/yr.  Add in support and
sales.  I'm skimpy, so that's $50K, and call the total about $225K/ year.

  In real life we provide a free molecular visualization and analysis
package (VMD).  Of the >1,600 downloads, about 400 are users with 
roughly linear growth during the two years of release.  The code
took 3 years to develop.  Assuming these number would be the same for
a commercial program, and assume we would have the same number of users
if the software cost something (balanced with lack of advertising, perhaps?)

 $225K/yr * 3yr = $675K, which we can round down to $600K since we
didn't have to support anything the first year.

  $675K / 400 users is $1688 / user, which just happens to be the same
price range ($1000 - $2000) you were describing.


  The big cost is the price of a programmer.  I assume you want good
programmers, and they cost (everyone wants good programmers).  About
the cheapest you could get would be about $30K/yr for salary plus
whatever is needed for insurance, benefits, etc.  This doesn't even
get you a factor of 2, so we're talking the cheapest might be about
$1000 for each sale.

  If this doesn't satisy you, what numbers did you expect?
BTW, I've also talked to other academic software institutions that do
market their product.  The cost is usually around $350 or so, and that
pretty much covers the overhead needed to market software.  Ie, a
usable interface, documentation, distribution costs, paperwork to track
users and inform of updates or bugs, etc, and support.  Again, price
it out and you'll see that that is pretty much the true cost of taking
any reasonable size project (> ~20,000 lines of code) outside the lab
in which it was developed.  They are cheaper because the cost of a grad
student is less than a programmer, and because usually the core part
of the code is the lab research, and is covered by grants.

  Again, if you don't agree with these numbers, show us a price break-
down.  If you think you can do better, start up your own company, or
better, be a consultant to these places and show them where they can
save money.

						Andrew
						dalke at ks.uiuc.edu




More information about the Bio-soft mailing list