3 ms·

Whats your target length(in hrs of effort) for a work sample? And secondly how do you deal with someone farming out the work to someone else?

tptacek 8y ago
Add up the time you'd spend interviewing on site and doing pointless phone interviews and that's the amount of time we allocate to ourselves for work-sample tests. In 10 years of doing this, I've never caught anyone cheating, and as I've remarked elsewhere and often, the quality of the hires we made once we switched to work-sample testing (a thing we did to raise the bar for our hiring, after our standard interview loop passed some people who were better talkers than appsec people) went drastically up. While I was at Matasano, we didn't let a single person go who had gone through our work-sample process. That said, if you're worried about cheating, build a small, non-confrontational interview into your process where you just talk about how they approached the challenges.
- JamesBarney 8y ago
  Thanks for answering, I was curious about the time because I know large tech companies will fly people in for day long interviews. But I've only been on the receiving end of a couple of interviews(funnily enough the last one was by sctb before he left for ycombinator) and they were all less than 3 hrs of my time from resume to offer. And I was worried a longer process might turn off candidates unless we have strong brand recognition, or an extremely competitive salary.
  - tptacek 8y ago
    It absolutely does turn off some people. And, what's worse/better: we refuse to alter the process for anyone; there are no fast-paths through our process. Running someone we're almost certain is "overqualified" through our challenges helps true up the challenges and the scoring rubric. I'm sure we lose people who can generally talk their way through any interview. And a lot of those people are good at delivery! But we gain so many more candidates we never would have been able to engage before that we come out way ahead in the bargain.

tptacek 8y ago

Add up the time you'd spend interviewing on site and doing pointless phone interviews and that's the amount of time we allocate to ourselves for work-sample tests. In 10 years of doing this, I've never caught anyone cheating, and as I've remarked elsewhere and often, the quality of the hires we made once we switched to work-sample testing (a thing we did to raise the bar for our hiring, after our standard interview loop passed some people who were better talkers than appsec people) went drastically up. While I was at Matasano, we didn't let a single person go who had gone through our work-sample process. That said, if you're worried about cheating, build a small, non-confrontational interview into your process where you just talk about how they approached the challenges.

JamesBarney 8y ago
Thanks for answering, I was curious about the time because I know large tech companies will fly people in for day long interviews. But I've only been on the receiving end of a couple of interviews(funnily enough the last one was by sctb before he left for ycombinator) and they were all less than 3 hrs of my time from resume to offer. And I was worried a longer process might turn off candidates unless we have strong brand recognition, or an extremely competitive salary.
- tptacek 8y ago
  It absolutely does turn off some people. And, what's worse/better: we refuse to alter the process for anyone; there are no fast-paths through our process. Running someone we're almost certain is "overqualified" through our challenges helps true up the challenges and the scoring rubric. I'm sure we lose people who can generally talk their way through any interview. And a lot of those people are good at delivery! But we gain so many more candidates we never would have been able to engage before that we come out way ahead in the bargain.