Michael, The size and scope of the project have a very direct relationship on the ability to cost the project -- and of course the bigger the project, the bigger the risk of a bad estimate.
I'll second the motion that requirements/design be done at either an hourly/daily rate, a separate project, or at least a decent flat rate proportionate to the size of the project. For a small project (few days), the requirements phase may be a significant portion of the time; for a large project, maybe only 10-20%. And of course development methodology and required documentation play a huge roll. Let's say you use Fusebox -- an hourly/daily rate for design and then a per-fuse charge make for a fairly reliable way to estimate costs. This also lets you cost out improvements/changes easily ("Well, adding those features requires about 10 fuses, so at $100/fuse, I can add the feature for $1000"). Databases can also be designed in a similar way, though I'd never tell a client that the database is "$x/table" -- but if you've designed databases, you've got a pretty good idea of how many objects, how complex the relationships are, etc, from the initial discussions. Building in a cushion for inevitable changes makes sense, but I find that 25% is pretty reliably keeping me out of debt For (web) design, I determine the number of templates (visual templates, not ColdFusion templates) and count the designer cost for that many pages -- which I tend to sub out. There's a basic cost for look/feel and then each major template would be an addon. It's crucial to build in maintenance costs, which can easily be the majority of your direct costs. I like doubling my initial estimate to account for maintenance and inevitable little tweaks and problems. It's also useful to remember that if you've doing modular development, you probably have a well-used set of application template components (eg security, menu system, basic template/layout, etc) that become fixed costs for you after a couple projects -- so if you lose a little on a project designing the initial security system, but it's one that will work (at least as a solid foundation) for other projects, you'll win over time. Regards, John Paul Ashenfelter CTO/Transitionpoint [EMAIL PROTECTED] ----- Original Message ----- From: "Michael Kear" <[EMAIL PROTECTED]> To: "CF-Talk" <[EMAIL PROTECTED]> Sent: Monday, January 06, 2003 7:19 AM Subject: How do you estimate development time? > I have a question that probably gnaws at most development houses at some > stage - what is the process you use to estimate the number of hours > you're going to need to complete a project? > > > Do you use a method that boils down to "we had a similar job to that > last month and that took us 55 hours and this job has that extra tricky > bit so it'll probably take us 58 hours" or are you more scientific than > that? > > > I have to change my methodology for billing my major client from a > hours-worked basis to a job-by-job basis which means I'm going to have > to be more accurate in estimating how much a job is going to cost me. > > > Cheers, > Michael Kear > Windsor, NSW, Australia > AFP Webworks. > > > > > > > > > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~| Archives: http://www.houseoffusion.com/cf_lists/index.cfm?forumid=4 Subscription: http://www.houseoffusion.com/cf_lists/index.cfm?method=subscribe&forumid=4 FAQ: http://www.thenetprofits.co.uk/coldfusion/faq Signup for the Fusion Authority news alert and keep up with the latest news in ColdFusion and related topics. http://www.fusionauthority.com/signup.cfm Unsubscribe: http://www.houseoffusion.com/cf_lists/unsubscribe.cfm?user=89.70.4