McFilter

Motivate the highest performers

From Joel On Software comes a link to this fascinating article about managing software engineers.

The main theorem is that good software engineers are many times more productive than even average engineers, and so the best approach is to motivate your good engineers and try to convert the average engineers into good engineers. Sounds like common sense in that previous sentence, but the natural tendency is to focus on the poorer performing engineers and "ignore" the better performers because they "aren't a problem".

Update:
Now that I've finished reading the article... I'm undecided as to how much I agree with the philosophy. A lot of it is geared towards getting engineers to work really long hours (or seems to be at least).

A long hour culture doesn't have to be oppressive - at STNC there was a good social atmosphere amongst the staff, so the long hours thing was (mostly ;-) fun. But then, we were generally young, and didn't know many people outside of the office, so we tended to hang out together regardless of whether we were in the office or out of it, so longer hours didn't mean neglecting other interests. I don't think that would work now that most of my friends aren't from in the office, especially in the summer when I'm playing sport at 6pm most evenings.

That said, I do feel that we achieved more at STNC than where I am now, but that could easily be through a lack of technical leadership, and/or my lack of commitment and ownership because I've not been here since it was half-a-dozen people.