Hi,

Welcome to the world of freelance consulting.  I'm sure you will get a
wide variety of suggestions here. Here is my fairly lengthy dissertation
which some may find controversial.

There are a couple of rule of thumbs that you need to use to determine
the cost for doing custom projects.  One industry standard for employees
is that they cost between 2 and 4 times what you are paying them for
salary.  The high end is for larger companies, the low end for small
ones.

For example,  if your salary expectations are $25/hour your employer
needs to budget at least $50/hour as his true cost of keeping you.
(Taxes, holidays, pension, sick leave etc.  His costs also include the
space he needs to rent/own and the cost of the physical stuff you need
to do your job.  Desk, Chair, PC, software, training.  I personally
believe that in this high tech electronics/software industry that 20% of
your time per year has to be allocated to training and upgrading.  PCs
have about a 3 year life and sometimes the software even less.

Cost of software packages is a very difficult one to access.  The
outright cost of a simple package might be low but the capabilities of
that software mean that you spend an extra 5 hours per week doing what
is trivial in a more expensive package.  At $50/hr cost that's $250 per
week.  It doesn't take long to justify a high end package at that cost.

Firmware generators are the next issue.  You've probably seen the
arguments on this list for assembler verses C for PICs.  That argument
has been raging for decades but the statistics are now fairly clear and
demonstrated by all research in this area that a good programmer can
produce between 10 and 20 '_documented_', '_debugged_' lines of code per
day and the language tends to not matter.

So if, like some of the list members you have a decade worth of assembly
language libraries that do everything including slicing bread,  the code
might be quite FORTH like with 10 lines of calls to 500 line library
functions.  If you are starting out,  10 lines of C code can do a fair
amount compared with 10 lines of assembler.

If the PIC you choose has 8K flash and your application fills 5K of it
any old C compiler will do the job and speed sensitive stuff you could
code in assembler.  But if that project expands to be 8K with the $100
compiler, and you spend the next two weeks optimizing to fit it all into
that 8K, you've more than exceeded the price of the high end expensive
tools that generate 10% to 30% smaller code size.

Test equipment and rent are two other costs that need to be factored
into your engineering rates.

So I think your guess at $25 per hour is very very very low.

Now, on to the development of a product for someone.  Once again the
hidden costs of manufacturing need to be factored into the production
costs.  For example, you need a switching regulator and all the parts
need to be surface mount for size constraints.  The Digikey prices in
100's mean that the price of the product per unit returns you $1 profit.
Buy the stuff in 5K reels and your profit jumps to $30 per unit but your
cost of stocking all this inventory goes from $1000 to $23,000.  (I'm
making up all these numbers).

You've just discovered another expense and that's the interest on the
money tied up in inventory.  If your client only buys 100 at a time and
ends up buying only 100 per year,  you are losing money big time.  You
may need a commitment from your client for 2000 pieces per year to
justify the cost he wants to get.  If you are hand assembling the boards
you'll find your hourly salary needs to be in keeping what they pay
assemblers in the third world to be competitive.  If you go automated,
there are setup costs for each run.  So it may pay to build more and
stockpile than to do small runs.

What I'm suggesting is do a business plan.  If you are not sure how to
go about doing that I can suggest a couple of avenues.  You could try
and find a student doing an MBA.  Part of the MBA program is to do the
exact sort of cost/benefit analysis you need to do to determine if the
project is worthwhile.

Alternatively,  and this may be a very effective method, contact your
local Chamber of Commerce, and see if they have retired businessmen who
consult or donate their time to help start-up companies like yourself.
A mentoring program so to speak.  And do listen to them.  Bill Gates
didn't get rich by being Mr. Nice Guy;   he applied sound management
principals to much of what he did.

So remove yourself from the personal issues of being a good guy or
developing a reputation so you sell yourself short.  The only reputation
you may get is that you are someone who works cheap while others reap
the benefits.

Hope that helps,

John Dammeyer





Wireless CAN with the CANRF module.
http://www.autoartisans.com/documents/canrf_prod_announcement.pdf
Automation Artisans Inc.
Ph. 1 250 544 4950


> -----Original Message-----
> From: pic microcontroller discussion list
> [mailto:PICLIST@MITVMA.MIT.EDU] On Behalf Of Jesse Lackey
> Sent: Sunday, June 16, 2002 6:39 PM
> To: PICLIST@MITVMA.MIT.EDU
> Subject: [OT] How to price freelance work...
>
>
> Hello all,
>
> I'm attempting to switch careers from Windows C++
> applications on salary
> to freelance electronics design.  This is both out of necessity (very
> bad job market) and personal passions (I love electronics +
> have a BSEE
> never used professionally and I hate microsoft).
>
> Well I have a client who wants a bid for a project.  The design is
> straightforward, it is half-clone & half-modernization/update of an
> existing product that I'm very familiar with.  Sorry for the vagueness
> but there is an NDA involved.
>
> Now he needs a revamp of this thing for use with other stuff in a
> "packaged kit" that he will sell.  This is a critical piece of the kit
> and there is no reasonable currently-existing alternative.  He has a
> small line of various products (that he doesn't manufacture, he's the
> retailer) into a niche market for years and knows his
> audience well.  He
> expects to sell 500 to 1000 of these kits a year.  He does not want to
> do any manufacturing.  I would be delivering finished product
> to him as
> his orders come in.  I also own the IP and can sell to other retailers
> or whomever else I'd like to.  He would get "preferential
> pricing".  I'm
> not sure how many I could sell elsewhere a year, 200 @ $100 markup ?
>
> A rough guess (detailed guess TBD very soon) is 3 weeks for the
> electronics design / pcb layout work to get to finished
> prototype, then
> whatever additional time for things like coming up with an enclosure,
> FCC issues, manual, etc.
>
> I will need to buy a schematic capture & pcb layout program.  Using
> xcircuit and pcb on linux will be too laborious.  It will probably be
> the $200 version of Eagle, so $400 total.  There will be
> chips and other
> parts to order, 2-layer pcb prototyping run, some SMD parts
> to solder, etc.
>
> I guess I'm looking for some rules of thumb for all this.  I think
> $20-$25 an hour is a good figure, I'd like to get more but this is the
> first "real" freelance project and I need to build a client base that
> can give solid recommendations.  However things are not so
> straightforward since I will be making money for each unit he sells
> (maybe $50-$75) you could look at it like he shouldn't have to pay me
> much at all, because over time I'll be making decent money
> (hopefully if
> his per-year estimates are reasonable - they seem fine to me).  On the
> other hand, he doesn't have to nor want to deal with
> manufacturing them
> 100 at a time.  That's my job.  And who knows, maybe this kit
> will be a
> flop and he only sells 50 a year.
>
> My current thought is budget $500 for prototype devel costs,
> absorb the
> cost for Eagle, have him agree to cover whatever FCC
> certification costs
> there are TBD, and another $3000 for my 6 weeks of time (i.e.
> about $12
> an hour).  Am I being too cheap?  My rent alone (we are both
> in the new
> york city area) is $1300/mo...
>
> I have a problem where I want everyone to like me and so I'm
> hesitant to
> put in a higher bid that might cause some disgruntlement
> towards me.  I
> suppose I have to get over this, since its business.  How
> much room for
> negotiation is customary?  (the aforementioned prices leave almost no
> room for me to go lower).  Is it acceptable to bid one price but agree
> to a counteroffer from the client?  Does this look bad / sleazy?  The
> "digital analogy" of submitting a bid, take it or leave it,
> if leave it
> there are no hard feelings has a strong appeal, but the world
> is analog
> and everything is negotiable.  What is the typical process for this in
> freelance electronics work?
>
> Ok this is quite a bit longer that I'd like but I wanted to
> get all the
> facts in place.  The piclist is a great resource and I hope to
> contribute back my experiences and knowledge once both have happened.
> :)  Any and all comments appreciated.  Thanks in advance everyone.
>
> Jesse
>
> --
> http://www.piclist.com hint: PICList Posts must start with ONE topic:
> [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads
>
>
>
>

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics