note for my friends: this post is targeted at companies and engineering managers. i know you know that hiring sucks and companies waste your time. this is a business case for why they shouldn't do that.
hiring sucks
most companies suck at hiring. they waste everyone’s time (i once had a 9-round interview pipeline!), they chase the trendiest programmers, and they can’t even tell programmers apart from an LLM. in short, they are not playing moneyball.
things are bad for interviewees too. some of the best programmers i know (think people maintaining the rust compiler) can’t get jobs because they interview poorly under stress. one with 4 years of Haskell experience and 2 years of Rust experience was labeled as “non-technical” by a recruiter. and of course, companies repeatedly ghost people for weeks or months about whether they actually got a job.
this post explores why hiring is hard, how existing approaches fail, and what a better approach could look like. my goal, of course, is to get my friends hired. reach out to me if you like the ideas here.
what makes a good interview
before i start talking about my preferred approach, let’s start by establishing some (hopefully uncontroversial) principles.
interviews should:
be able to tell the difference between a senior programmer and a marketer using chatgpt. reflect the actual job duties. this includes coding. but it also includes architecture design, PR review, documentation, on and on and on. all good senior software engineers are generalists. select for applicants who will be good employees for years to come, not just in the next quarter. people are not fungible.
there is a high cost to losing employees who are a good fit to the project.
... continue reading