I’ve hired 450 Upworkers over the last eight years and paid $4.2M in wages.  I think that makes me an Upwork superfan.  

Interviewing, whether on or off Upwork, pains me.  I do my best to ask rigorous questions and score them consistently, but I know the data: I suck at interviewing and so do you.  We all do.  Decades of studies consistently show that interview scores are only weakly correlated with subsequent job performance.   

I’m too jaded to be surprised when a panel of five interviewers unanimously vote “Strong Yes” only to have the candidate underperform as soon as they're onboarded.

Google’s old SVP of People, Lazlo Bock thinks there is something better than interviews:

The best predictor of how someone will perform in a job is a work sample test. Give candidates a sample piece of work, similar to that which they would do in the job, and assess their performance at it.”

But running full-day work samples with candidates is pain in the ass and expensive, so we all compromise.  A quick 1-hour live exercise or some sort of take-home project is better than nothing, we rationalize.  We give up our wish to really see candidates in action for the sake of expediency.

What if, instead, we could work with three finalists for a week and then pick the one to hire?  How much better would our hiring decisions be?

That’s the Upwork Rule of Three: hire three people for every one opening.

When I hire on Upwork, I pitch the first gig like this: “We’re looking for someone to help us for about 20 hours to try us on for size.  If it’s a great mutual fit we’ll turn it into a long-term collaboration.”

I use all the usual selection techniques to whittle down my shortlist to a top three and then hire all of them.  If possible, I give them all the exact same assignment: build this feature, write this article, or design this page.

After a week of working together I thank two of the three Upworkers, pay them, and end the contract. There’s no hard feelings -- they got paid and earned a positive review.  And I built a deep, accurate perspective on what it would look like to work with them long term.

In a perfect world, I’d have conventional hires go through a similar paid trial period before I gave them a permanent offer -- what better way for both of us to figure out if we’re the right long term partners?

Doesn’t it seem crazy that the vast majority of us make life-altering career choices without ever having worked together?

That's why we replaced the conventional final round interview day with a collaborative project day, so we get to work side by side on something meaty.