On the TechTO stage ×1
First seen on a TechTO stage in 2016. Every TechTO talk is searchable — ask the archive about Farhan ↗
In their words
Interviews are a horrible predictor of performance — so guess what? I didn't do them.
They said, 'Well, we have this very strict interview process where only 1% of candidates get from this funnel to this thing.' And I'm like — but what was the goal? The goal for us was to hire people. So we hired them at the event.
I interviewed at Amazon — it was eight hours of interviews, and eight hours tell them nothing about how good I am ... Internships are a very, very good interview.
Quick answers
How fast did Xtreme Labs grow?
From 10 to 350 people between 2009 and 2013, hiring almost a thousand people in Toronto over four years — with a team of roughly 35% women and 25% female engineers. In one Waterloo round they hired 65 engineers, beating out Facebook, Twitter and Google.
How did his on-the-spot hiring work?
At a University of Toronto career fair, with no booth materials, he gave candidates a short programming problem, discussed their answers, then said 'you're hired' — seven hires in one day, all starting Monday, none asking about salary. He made every decision himself, with no hiring committee.
Didn't instant hiring produce bad hires?
He says attrition was only about 1% after 90 days. His logic: interviews poorly predict performance, so compress them into about 15 minutes and evaluate people in the job over 30, 60 and 90 days — internships, he notes, are a very good interview.