Posted by lawrence k on 09/19/07 20:33

On Sep 18, 8:27 pm, Sanders Kaufman <bu...@kaufman.net> wrote:
> So now, these PHP recruiters are calling out of the blue and from
> everywhere. My phone ain't rung like this since the Clinton years.
> Let's say I'm getting $10/hr - and man, it's sooo much more than that! -
> but let's say. If I do it the way they expect it to be done, I can milk
> the contract for what - 8x40hr weeks - that's $3k plus. But if I do it
> in a 16 hr weekend - I don't even get $200.
>
> I was thinking that, when I talk to them next, I should suggest a
> project-based, rather than hourly-based, rate - and estimate at the high
> end. But I've never done that before through a recruiter, and I don't
> know how it'll go over. I just know I've *always* had problems with
> hourly recruiter jobs.


I used to work at category4.com, and they did all their projects on a
per-hour basis, but they became upset by the same conundrum you face -
when they got done quickly and under-budget, they were penalized by
getting less money. So they've switched over and now do all projects
to a flat fee. However, estimating the right amount is a total
nightmare. It's also difficult to ensure that you and the client have
agreed to the same feature set. Misunderstandings are common. I wrote
of some of the pitfalls I've seen:

http://www.teamlalala.com/06-05-07_marketing.htm

One idea I've warmed up to is "Start with the interface first".
37Signals.com has been pushing this idea hard. For big projects, I
insist the client hire a designer and design the whole thing before
I'm willing to start writing code. The goal is to reduce risk (and
financial loss) by working out as many details out as possible, before
the first line of code is written.

Despite that, it is still common to under-estimate projects.
RawStory.com recently asked us to rebuild the code behind their site.
We estimated $50,000, which they thought was too high. Yet we used
multiple methods to work out what was involved, and each method we
used steered us back to a similar answer (give or take $10,000). My co-
worker is certain that at $50,000 we under-bid the project. All the
same, RawStory.com went looking for someone else.

That brings up another problem: you've got to educate the client about
how much work is involved. This won't be possible if you yourself
don't know how much work is involved.

Category4 is the most successful web design firm in central Virginia.
I've often wondered why, and I've decided the probable answer is their
focus on client management. The client managers provide a buffer
between the client, on the one hand, and the designers and
programmers, on the other. That buffer is crucial - it helps stop
feature creep. The client managers goal is to provide the discipline
that keeps a project on time and on budget. Without that, it is
inevitable that a project will mutate while work is going on, and come
in late and over-budget.

[Back to original message]