It's time for data science to be part of your hiring process

Every company claims that data and hiring are crucial, but few companies use data in their hiring process


Nearly every CEO will tell you human talent is the reason their business is successful: A great sales hire can change the direction of your entire company, while a bad engineering hire could result in your product falling flat on its face. Nearly every CEO will tell you they are “data-driven.” So if you know data is important and the people you hire are important, why aren't you using data to hire?

I asked myself this very same question a few years back. What I decided to do was test the hypothesis by grading the hires I'd made over the past few years on a scale of “disaster” to “superstar.” Then I looked at some data around each of those employees and tried to tease out the patterns. Why were people successful? Why weren't they? And most importantly, how could I use what I learned to hire better in the future?

Some large companies do this already, by the way. Example: There's a reason Google offers bonuses to move closer to the office; it has found that employees who live closer are more likely to stay in the office, to do better work, and to stay at the company for longer. Which makes intuitive sense. Nothing sours a job you actually enjoy like spending a couple of hours getting there and back every day.

The lessons I learned from scoring my own hires might not be the same as what you'll find if you do likewise. But you'll find similarities and personas you probably couldn't articulate beforehand. Here are a few of the data points I recommend using to build out a hiring persona that's right for your company:

1) How much experience is enough?

Experience almost always correlates with success. But sometimes experience isn’t the single most important thing. Folks with tons of experience are more likely to get competing offers or find certain tasks beneath them; meanwhile, people who feel they have a great opportunity in front of them may work harder than anyone on their team.

2) Did they come from a fancy college?

Some companies insist on hiring Ivy League graduates. But does the data really support this?

3) Did I hire them for a junior or senior role?

This will often dovetail a bit with the experience question, but what makes an employee in an entry-level position effective might be totally different from someone in a senior role. This is, in other words, an important data point to use when you're looking for success factors.

4) Did they ace my reference check?

This turned out to be one of my most important data points. Keep in mind I'm not using employee-provided references since, well, if your prospective hire can't find two or three people to vouch for their skills, they probably aren't that good in the first place. I'm talking about looking through their experience and checking with friends who worked with them at their previous companies. These are people I personally trust and who I know wouldn't bend the truth or exaggerate someone's bona fides. If they pass my own reference check, they usually excelled at CrowdFlower.

5) How strong was their referral?

Here’s where I really found a strong correlation to employee success. When I dug into the data for my hires, the employees I rated as “disasters” were never employees I was referred. That’s not to say that all our best employees were referrals, just that none of our disastrous ones were.

Meanwhile, the strength of our worst employees’ alma mater, however, was actually higher than average employees. On the other hand, the best employees tended to come from fancier schools.

employee outcomes Lukas Biewald
Employee outcomes

Conversely, you can see that referral strength rose in tandem with employee quality. In other words: If you came to CrowdFlower with a great referral, odds are, you’d be a great hire.

So how can you use data on your hires? It's fairly simple. Have you and a few of your execs grade your ex-employees. Focus on a single, simple question: Would I hire them again? Then, start looking at some of the criteria I listed above and score each. A few others you can think about using:

  • Did they come from a startup or bigger company?
  • Did they major in what they actually do?
  • Are they a generalist or an expert?
  • How far away did they live?
  • Did they know someone else at the company before they started?

Once you're done, do what data scientists do with good, rich data: Find patterns. Find what worked for your company. Data alone can’t tell you who will be a great employee, but reflecting on who you hired and how they did is incredibly important. Even if it’s a small data set, the patterns will help you make better hiring decisions, and making better hiring decisions will make your company more successful.


Copyright © 2015 IDG Communications, Inc.

It’s time to break the ChatGPT habit
Shop Tech Products at Amazon