|
|
Posted by The Natural Philosopher on 09/19/07 11:27
Sanders Kaufman wrote:
> The Natural Philosopher wrote:
>> Sanders Kaufman wrote:
>
>>> *The problem is that I can build the damn thing in a weekend, but I
>>> don't want to miss out on all that extra cash.*
>>
>> You may be able to build it in a weekend, but what about all the
>> documentation for the NEXT guy who has to modify it, and the user
>> documentation, and the training course and manual you have to write.
>
> Well, there's an idea.
> I do love to fill binders.
>
> I just don't want to give them too much more than they want, lest the
> application exceeds their needs by an order of cro-magnitude.
>
>
>
>
>> Stop thinking like a programmer, and start thinking like a doctor.
>>
>> They don't want CODE. They want a clearly described simple SYSTEM to
>> do a specific JOB. You job is not to deliver code, but a complete SYSTEM.
>
> That's what my post is about. I'm not sure *what* they want or if they
> know what they want, or if I can convince them that I can give them what
> they want, but that they don't know they want yet.
>
Well go in there and ASK them.
"I am a programmer and system documentation writer. I haven't a clue
what you want, but whatever it is I can write it. Our first priority is
an exchange of knowledge: I need to understand what you want and you
need to understand what is impossible, so we don't try and attempt it:
then my first job will be to document this conversation in a short
preliminary specification that you can read, and we can argue over till
its possible, what you need ,(and even maybe what you want) . This
process will take longer that writing the code (which is not hard), and
about as long as writing the manual, : What IS hard is for me to
understand the nature of your particular problem, and for that I need
your time and assistance"
> I just know I want the gig.
>
> (OMG - I think that's a direct quote from Pirates of Silicon Valley!)
>
>
>> Preferably with 30 nicely bound user manuals, and so on.
>
> I think that's my answer right there - although I think one or two
> binders with some very wordy help screens and wizards might do the trick.
Ask them. Write a spec. Be professional. The last $500k job I
did..probably the last in my professional career - I spent about 4 days
installing and setting up kit onsite, I spent 6 days configuring at
base, I spent 2 weeks writing the spec, and another 5 days writing
training courses and 14 days delivering them.
We spent more time making the sale than implementing the solution.
What people pay good money for is complete turnkey SOLUTIONS to their
PROBLEMS.
Thats is FAR more than delivering code that works. Make sure they
understand that. You are first and foremost a business analyst, with
technical knowledge. Secondly you area systems analyst, and finally your
are an ANAL PROG (as we called) them coding to a spec, and after that
the dreaded TECH AUTH and CODE TESTER who rip the thing to pieces and
find all its problems and document the thing in PLAIN ENGLISH for the
people who think computers are spawn of the devil..
The ANAL PROG bit is actually - if the spec is clean and well founded, a
piece of piss.
The rest is NOT.
I am as a favour for some mates, busy coding up a back office system. I
have spent nearly as much time with them watching how they work, and
understanding where a computer might speed things up (surprising how
often it slows things down tho if you get it too clever) as actually
really coding it.
They are completely awestruck by the fact that I am asking THEM how they
do *their* jobs..and regarding them as the experts they are in doing it.
Frankly coding it all up is a deeply uninteresting chore: The interest
for me is analysing the business requirements into a set of computable
concepts etc. etc.
Navigation:
[Reply to this message]
|