> I'm not really looking for sympathy... just opinions.  I realize now
> that I should have stated my questions more definitively.

The real answer to your question is that an acceptable failure rate
is one that doesn't cause your customer to comment.
If he's come up with a magic number of 1%, then your worst case
target is less than that.

What you can't determine (although the marketroids will extract a
number from their belly-button) is the loss of sales from lack of
confidence.

What you can determine is what failures are worth to you.
Take a hypothetical case where you are making 1000 units/day and you
make $100 on each one. If the like expectancy of your product is 3
years, you can expect to be getting 10 units/day back for servicing.
Lets say that the average cost of repair and freight is $50. You'd
also have to add in the overhead of the staff, equipment and space to
have a service department. That might be $25k/year or about $100
per working day.
So the cost of a 1% failure rate is $50 * 10 units + $100 = $600/day
or the profit on 6 units. What would your boss say if your early
morning ritual was to have a cup of coffee and then go down to the
factory and take a sledge-hammer to 6 units ?
An alternative approach is that this is about $150k per year. Go in
and ask the boss if you can have half of the money you save added to
your salary.

Steve.
======================================================
Steve Baldwin                Electronic Product Design
TLA Microsystems Ltd         Microcontroller Specialists
PO Box 15-680, New Lynn      http://www.tla.co.nz
Auckland, New Zealand        ph  +64 9 820-2221
email: steveb@tla.co.nz      fax +64 9 820-1929
======================================================