Make precise estimates and meet them

Working with the unknown – entirely new techniques, new technology, new ways of thinking – makes it hard to estimate correctly. I’m missing by far, on the side of being too optimistic. Bah.

Some specific techniques that I use:

Estimating Accurately
The trick to this is to always aim-privately-for actually having the project done earlier than you promise publicly. I even use this approach to always be on time for appointments. If I need to be there at 6:00 and it’s a 30 minute drive, I plan instead to be there at 5:45 and I really try to be there at 5:45. I haven’t been late to anything in a year—when I’m “late”, I’m actually on-time. Same goes for projects: always under-promise and over-deliver. If they think your estimate is too far out, then they’re not ready to work with someone with your level of integrity. Get burned by a few people who use “over-promise and under-deliver”, and they’ll come back to you every time.

The Perfect is the Enemy of the Good
Perfectionism is often a challenge when you under-promise. The theory goes that you’ll think “Oh, I have time” and put it off. In reality, if I find myself procrastinating a project, it’s a good bet that what’s really going on is that I’m planning to do a perfect job. What ends up happening is I wait until the last minute, then I do a last minute job instead. To combat this, I’ve begun doing the last minute job first. Then, one of two things happens: either I’m so motivated to do a better job that I have no trouble with procrastination, or I get the “last minute” version done and realize that it’s good enough.

I recently realized that I’ve successfully accomplished this goal. I’m to the point where I can take any task given to me by someone, break it down roughly into the steps involved, and tack on a time estimate which I am always able to easily meet.

To say it has been “worth doing” to master this is an understatement. My pride in the quality of my projects has gone through the roof. I feel like I’ve broken through to a new level of professionalism that many of my colleagues still struggle with—and I’m now comfortable charging substantially higher prices.

Customers, clients and bosses now return to me for their projects instead of going to the guys who charge less and promise the world. After being burned by one or two of those guys, my approach suddenly has an irresistable appeal.

Less vague, more useful. This is something I want to do.

It seems I keep missing deadlines I set myself. I’ve been wondering why this is, and come up with a couple of reasons:

I’m almost always treading new ground
This means that I’ve no or very little experience to draw on when I’m making my estimates.

I feel pressured to estimate lower than I think realistic
Wishing to please I will say two to three days when my subconscious is screaming that a week is more likely.

I often don’t have time to make a qualified guess
I’m often asked to make ball-park estimates off of the top of my head, which makes me rush through and reach a result where I haven’t calculated with everything that needs to be done.

I procrastinate
I usually jump into a task working with great focus for the first while. Then, having made good progress, and knowing I don’t have to deliver until next week, I feel I can spend a couple of days just fiddling around, getting distracted and such. Problem is, the feeling doesn’t go away as the deadline nears.

And finally:
I get away with it on a regular basis
I’ve yet to be even reprimanded for missing a deadline. Not at school, nor at work has anyone scolded me for missing an assignment. Learning that I could get away with it, I prioritized my time differently.

Wanting to change this, I think I’ll have to work on each of these points.

Treading new ground is in my circle of concern, but not so much in my circle of influence, which means I can’t change the fact that I’m often working with new things. It’s what I love doing, even. I can, however, stall the estimate until I know some more about the technology and the demands.

Trying to please by estimating low is in my circle of influence. I have to remember that when I say something takes a week, it’s going to take a week no matter how much you need it tomorrow, and I’m not really pleasing anyone by not delivering when they expect it. Plan with the end in mind, I guess.

The having time issue is partly in my circle of influence. I can refuse to do ball-park, which isn’t going to work, or I can calculate with a buffer of sorts, something which I should probably do in all my estimates.

Procrastination is totally in my circle of influence. Nothing outside of me can change that. I have to work on doing first things first, getting important things our of the way before they become both important and urgent.

Getting away with it is only a problem as long as all the other things are. If I don’t underestimate my tasks, and always deliver on time, I won’t need the habit of getting away with missing deadlines.

... my duty to improve on my programming. I know that I am capable of doing crazy wicked things with a computer, but I also know that I have to get better at structuring my work, and create less errors.

My main focus for now is to get into unit testing, and finally grasping how to test everything – I mean, I understand how I can check that a function returns the expected value, but how do I easily test whether a ‘create new user’ function works, including whether or not the email sent out is correct, and such. I guess ruby on rails has something of an answer to that in having a development and a test database… hm. Another example is some of my work code where I generate PDF files. How can I test that the right text gets put in the right places?