Hacker News Reader

Comment by tptacek | Hacker News Reader

I agree strongly. Contract-to-perm hiring seems like a great way to filter out the best candidates, who will tend to have better options.

View on HN

It's a pretty big issue because you don't really know what working with them will be like. Weebly does a trial week, which is a good balance of time and commitment. http://www.sequoiacap.com/grove/posts/akzj/trial-week-our-hi...

Aside: let's be careful not to pretend that temp-to-perm hiring is any kind of norm. The overwhelming majority of hires are hired conventionally, after an interview, with an immediate full-time offer. That includes "employee #1" hires.

You're right

I'm also a crusader on this issue, as you know, so don't let me steamroll the thread.

You've made this point before and it seems like a good one, but it raises a question. If you don't have a trial period, does that not make it harder to correct mistakes? The barrier to quickly firing someone who (perhaps through no fault of their own) turns out to have been a bad choice is surely higher than allowing a trial to expire or whatnot.

Since bad hires are one of the worst things that can happen and no one is perfect at hiring, I'm curious what you do about this problem. Do you just work harder at getting closer to perfect?

It seems to me that there's always an implicit trial period anyway—the question is how you encode it. The best people, who have many options, won't want to stick around if they made a mistake either, right?

Apart from adverse selection and basic fairness, which I think are the big issues with temp-to-perm hiring, I also don't believe the practice does much to mitigate risk. Think about the kinds of things that demonstrate a teammate's unsuitability for the team; how many of them are consistently apparent in a 1 week sprint? Anyone can be productive for a week. Anyone can avoid teamicidal petulance for a week. Anyone can avoid making catastrophic design choices that won't be apparent until after you ship for... well, &c &c.

Reply on news.ycombinator.com