4 ms·
> avoid project-based estimation You'll have happier clients as well, because with experience, the buffer in project-based estimations needs to be large enough
by groestl 8y ago
> avoid project-based estimation
You'll have happier clients as well, because with experience, the buffer in project-based estimations needs to be large enough that no possible hidden complexity can deplete it (barring complete show stoppers). Clients will usually be better off with time-based billing when the per-project cost estimate would come from an experienced provider.
- hardwaresofton 8y agoOK, so then in my head the advice is now: "do time-based billing but avoid hourly increments, instead opting for weekly/monthly but never disclose how many hours per week you work". It seems inevitable that clients wonder/ask why you can't get X, Y, and Z done in a "week", when they hold a definition of what a "week" means to them. If you're billing for a week, they're going to assume you're billing for 5x8 hour days -- is the idea just to let them assume that, but you actually work however many hours it takes to get the deliverables for the week done? If so, that's basically forcing you to over estimate, or work a full work-week anyway, and then you might as well be pricing by hour (whether you show them or not). What I've done recently is charge for X days/week, with a written agreement specifying what constitutes a "work day" for me. Clients normally don't care/micro-manage (there's enough mutual trust), and when we meet to show/discuss progress, I let them know how how "on schedule" we are (so at "based on how quickly I'm working at the current clip of 3 days/week, it looks like everything will be done in 4 weeks"). This formula would be wrong according to the usual HN advice though -- it's way too easy to translate X days/week w/ a set amount of hours that constitute a work day to X-per-hour billing.
- ascar 8y agoI think the important point is not that it can't be broken down to a hourly rate, but that the smallest incrememt is 1 week or a day. They simply can't book you by the hour. Specifying what a week or day is, is up to you. In regards to time as well as money.
- lmm 8y ago> What I've done recently is charge for X days/week, with a written agreement specifying what constitutes a "work day" for me. Clients normally don't care/micro-manage (there's enough mutual trust), and when we meet to show/discuss progress, I let them know how how "on schedule" we are (so at "based on how quickly I'm working at the current clip of 3 days/week, it looks like everything will be done in 4 weeks"). This formula would be wrong according to the usual HN advice though -- it's way too easy to translate X days/week w/ a set amount of hours that constitute a work day to X-per-hour billing. Why do you need that "written agreement specifying what constitutes a "work day" for me."? That sounds like it's just encouraging the client to micromanage; you don't want a client that doesn't trust that you'll do a decent week's work in a billed week. And it gives you the disadvantages of a conventional job: you're now tied to working X hours/day Y days/week rather than taking a long lunch some days, working late some days, stopping early some days, and trusting that it'll all average out.
- hardwaresofton 8y agoI think that most clients in my current part of the world (Tokyo) have a set expectation of what a work day is -- normally ~8(+) hours. I have that in the agreement to avoid this misconception if a certain client expects me to be available/reachable during all of their own working hours. Also, I do maintain flexibility in how I work the hours -- I don't say when those hours will be worked, just how many "days" are being billed per week. Sometimes I split it up over several days, and I almost always split it up into multiple chunks on a single day. This is the part I like most about the arrangement and my current lifestyle. Eventually you have to do some work, and the number of hours/day is the number I'm comfortable with working (it's less than 8). Now this completely goes out the window for project based pricing jobs -- If I am charging for project/milestone completion, then I do not mention "work days" because they're irrelevant, there's only the deadline and whatever scheduled checkups/sync meetings, doesn't matter if I completely ignore the project for a whole week. This comes back to my point before -- it seems to be the only real dichotomy is whether you charge for 1 unit of time (before/after the project deadline) or whether you charge for some derivative of hourly wages (due to the preconception of what a "workday" means in most people's heads). Assuming you understood my previous point, are you suggesting that I should just leave the definition of "work day" vague? and treat a weekly contract as a project based one? I guess this could work, but it feels more to me like project based pricing split up into milestones that happen to have weekly delivery dates -- not what everyone is saying to "charge by week". And the problem there is the same -- how do you avoid/deal with bad estimation?
- lowercased 8y ago> It seems inevitable that clients wonder/ask why you can't get X, Y, and Z done in a "week", As much as I like the 'per week' idea, there's often stuff I can't get done because someone is sick during a week. Client's don't particularly like that. If they wanted to be paying someone $x/week regardless of work output, they'd hire more employees - they're often using external contractors for multiple reasons, but efficiency is usually one.