Hacker News Reader

Comment by lbj | Hacker News Reader

More generally I would say that you need to religiously manage expectations and to keep in mind that this is much easier going into a project than halfway through. Giving an estimate of "1 month" is easy. Giving "1 week" 4 times is very painful, which is why I always work off buffers.

View on HN

In a nice work environment giving an estimate of 1 month is easy, but not always. A lot of managers have the bad habit of trying to negotiate estimates and it takes quite a bit of confidence and even stubborness to avoid telling someone what they clearly want to hear.

That's the job but it can be one of the hardest parts of the job.

Then for the next project, you know the pushback you're going to receive so (perhaps unconsciously) you have this bias toward a lower estimate.

If you're not a regular employee the incentives are even worse. If one contractor tells you a week and the other tells you a month, which one will you hire? When you're billing for time and materials, and that's pretty normal, it hardly matters to the contractor if a deadline is missed because most clients know switching mid-project will only add to the delays.

Yeah negotiating estimates with employees is odd. You are trying to predict something. Negotiating is to skew the prediction to be less accurate on average. Employee is not even doing it for fixed price so all the negotiating down does is lose accuracy.

Some really shit jobs would expect weekend work to make the difference but with this job market: f that!

You’re not trying to make it more accurate, you’re trying to shift ownership of the resulting shambles to the individual employee.

Yes exactly. Or they just don't trust the employee, either ethically (they'll slack off) or technically (they are doing it wrong, that's why it takes long). Or worse they are of the "they need to be challenged, keep 'em out of the comfort zone" mentality.

Thinking about this, what if you prepare for the meeting and write down the estimate! What are they going to do, argue with a piece of paper? Scribble some notes to make it look like you were thinking, and write the estimate, then present it as your "original unbiased estimate" which, of course, will never change. This won't always be an option, but it's a good trick to have I think.

Contractors don't know how long projects will take.

I've given up on deadlines for projects; now we just have contributors work on small deliverables (that they usually define) that build toward the goal.

Not having deadlines makes learning about a new unknown unknown a lot less painful.

This is the smartest way to do it.

Reply on news.ycombinator.com