> I hate tech interviews, they’re a poor proxy for real work.
> Here are some ways they’re silly (some examples follow).
> A good interview would not have these bad features, maybe someone should do that.
What I’ve observed in 12 years of practice is —
It works; things go south quickly if you don’t do rigorous technical interviewing.
There is huge variance between interviewers, in both accuracy and bedside manner.
The most effective signal by far (close to 100%) is: someone I already trust technically who has worked with them says they’re good, and is willing to stake their reputation on it. I enthusiastically skip the more tedious parts of interviewing in these cases.
Interesting things I’ve seen, experienced, or practiced:
Long trial contracts to hire, no exceptions. Buys you accuracy at the cost of reduced candidate pool.
Willingness to fire quickly (first 30 days) — in practice similar to above. This is often advocated for but rarely implemented.
De-emphasize interview, heavily game-plan and weigh reference calls; do a lot of them (5+). This works if you have the technique to blow past the boilerplate endorsements.