3 ms·
Providing feedback is pretty hit or miss - if you give anything specific enough to be constructive you also end up with people often wanting to argue the partic
by itsdrewmiller 11y ago
Providing feedback is pretty hit or miss - if you give anything specific enough to be constructive you also end up with people often wanting to argue the particulars. The feedback examples here are pretty bad. "You don't have enough experience" - so you are going to hire him or her in two years once they have gained experience? "You aren't a culture fit" - why not? Is it because he or she isn't a 20-something white male who likes video games and beer?
The feedback I have tended to give is along the lines of "Our process has a high chance of providing false negatives, so just because we are not offering you a position doesn't mean we are convinced you are not qualified. On [Programming Problem X] we felt that your code involved too much deep nesting and the variable naming was poor; in an environment with many programmers that is one of the most important things to do well. You also missed [Requirement Y] which, while not critical for the functioning of the program, did show a lack of attention to detail of the problem statement." And then they still argue. :-)
If they are really horrible they get rejected in real time, and if it's late in the process we have a conversation internally about how they managed to make it so far.
- klenwell 11y agoI think your first line is a great generic response and sufficient in itself: Our process has a high chance of providing false negatives, so just because we are not offering you a position doesn't mean we are convinced you are not qualified. I don't really care for the reason why I'm rejected. I'm deluded enough in my own abilities to rationalize the reasons why you're making a huge mistake. :) The main thing I want is a timely response. What drives me crazy is not getting any response at all.
- itsdrewmiller 11y agoThat is something I didn't realize before but seems to be a common theme in the comments here - So I learned at least one thing to do better from this thread, even if it wasn't from the article. :-)