The first chat with the technical person should highlight if there is a need for a coding challenge at all. Like, the very strong candidates are so thoughtful that's essentially pointless to ask them to code. They are also rare and far between, though.
The average candidate can code and something can be said based on their coding challenge, which is not necessary something measurable or comparable, but it is something useful to understand who that candidate is.
I'm struggling to express my thought because I'm the first one to say "I should be able to chat with a person for ~X hours to know if I want to hire them", however I'm faced with the reality that:
- it's expensive (people hiring should be your strong people)
- guts are not always right, occasionally they fail and that candidate is not good
- the best experiences I had as an interviewee did involve some live coding, although that wasn't necessary the focus of the interview: getting to know the people on the other side and pair with them is what made it a good experience. I dread live coding, but in some form those interviews were great
- met candidates which were amazing in all forms of chat, everyone hyped, until we faced them with technical stuff (very basic)
I don't think there is a saving grace for the process I'm currently using, however I can't find a solution either. A simple chat with the team has proven dangerous as well as expensive multiple times.
An objective process has proven impossible and rigid. We are talking with humans, I'd rather tune every evaluation to the candidate itself, which in turn leads to making the process hard to evaluate.
Referrals work only up to a certain point, given the amount of people needed to hire, it's not a funnel big enough.
I wish I could solve the problem in some way, I read and study, but I still haven't found a way to solve it.
Almost forgot:
The coding challenge keeps being a useful way to sprout conversation during the "full day with the team" interview, so it sets the ground for that too.
The average candidate can code and something can be said based on their coding challenge, which is not necessary something measurable or comparable, but it is something useful to understand who that candidate is.
I'm struggling to express my thought because I'm the first one to say "I should be able to chat with a person for ~X hours to know if I want to hire them", however I'm faced with the reality that: - it's expensive (people hiring should be your strong people) - guts are not always right, occasionally they fail and that candidate is not good - the best experiences I had as an interviewee did involve some live coding, although that wasn't necessary the focus of the interview: getting to know the people on the other side and pair with them is what made it a good experience. I dread live coding, but in some form those interviews were great - met candidates which were amazing in all forms of chat, everyone hyped, until we faced them with technical stuff (very basic)
I don't think there is a saving grace for the process I'm currently using, however I can't find a solution either. A simple chat with the team has proven dangerous as well as expensive multiple times. An objective process has proven impossible and rigid. We are talking with humans, I'd rather tune every evaluation to the candidate itself, which in turn leads to making the process hard to evaluate.
Referrals work only up to a certain point, given the amount of people needed to hire, it's not a funnel big enough.
I wish I could solve the problem in some way, I read and study, but I still haven't found a way to solve it.
Almost forgot: The coding challenge keeps being a useful way to sprout conversation during the "full day with the team" interview, so it sets the ground for that too.