Digital Staffing: The Future of Recruitment-by-Algorithm

Americans are now spending more time on social networking sites than on all other sites combined. Facebook alone has more than 1 billion users — that's 15% of the world's population and almost 50% of internet users, and they spend an average 15 minutes a day on the site. And that's just one site; imagine if you added in Twitter, LinkedIn, Google+, Weibo, Renren, Orkut, and on down the list.

As a consequence of spending so much time online, we now leave traces of our personality everywhere. Indeed, unless you have never used Amazon, Gmail, Spotify, Tripadvisor,Netflix, or any multitude of other sites, you will have not one, but multiple online profiles. In the beginning, the profiles were of interest only to those websites, which customized our consumer experience by offering us products congruent with our preferences and values ("if you bought this movie, you may want to buy..."). However, our online behaviors are now also of interest to recruiters and employers, who are desperately trying to translate them into "digital reputations" and use them to find talent online. I see three reasons that employers are likely to find their future leaders in cyberspace.

First, the web makes recruiting easier for employers and would-be employees. For instance, a company with 100 employees will probably have close to 100 employees on Facebook or LinkedIn, and each of them will have at least 100 connections on these networking sites — this means targeting 10,000 people who are first-degree connections, and since they will have at least 100 connections each, the job ad could reach over 1,000,000 if we include second-degree connections. For employees, killing time on Facebook or Twitter while at work may not be that pointless after all — it can help you find a more desirable job (or be found). Indeed, 1 in 6 job seekers credits social media with helping them find a better job.

Second, the web makes recruiting less biased and less clubby. Most recruiters are already using social media to identify talented employees outside their usual networks. According to a 2012 survey by Jobsite, 54% of recruiters use Twitter, 66% Facebook, and a whopping 97% LinkedIn, as recruitment tools. While this widens the pool of recruitees, recruiters are still subject to the same biases that operate in the physical world (notably prejudiced inferences about someone's character or values based on their appearance). However, it is easier to create and implement reliable methods online than offline, where chemistry and subjectivity will never be eliminated. Conversely, digital reputations capture many hours of online behaviors, and unlike with stocks, with human beings past behavior is the best predictor of future behavior.

Third, web analytics can help recruiters become more efficient. Big data can provide the best answer to the big questions in talent identification, if we ask the right questions of the data. Not only is there an abundance of data, it is also getting easier, quicker, and cheaper to generate more (relevant) data. Data aggregation algorithms are growing exponentially — Klout may not be the best measure of "social influence," but it is still useful and fairly reliable, and future alternatives will no doubt be improvements. Data integration — combining people's multiple profiles into one — is the next step, and it's already happening. Soon, it will be easy to know that the person who buys Colin Dexter books on Amazon is the same person who streams Inspector Lewis on Netflix, checks out the Randolph Hotel TripAdvisor, and searches for flights to Heathrow on Kayak. If you've ever shopped for a pair of shoes on Zappos and then seen those same shoes advertised to you the next time you visit Facebook, this has already happened to you. The question becomes: how can recruiters put all these pieces together to quantify potential hires?

If you think this is scary, you may want to consider the alternatives: missing out on a better job, spending ages updating your CV, completing dozens of individual job applications, or living your life entirely offline (which would be a very lonely life indeed). Furthermore, being a luddite will probably damage your career. Recruiters will deem candidates unemployable if they fail to find information about them online — unless you are hiding an undesirable history or do not exist, you are now expected to have an online profile.

The big implication is that you need to invest a considerable amount of time managing your digital reputation. The only thing worse than not having a profile is having an undesirable profile. Indeed, your chances of being headhunted online are inversely related to the amount of inappropriate self-disclosure found in your Facebook or Twitter profile. Egosurfing — self-googling — is now more important than updating your CV.

We will soon witness the proliferation of machine learning systems that automatically match candidates to specific jobs and organizations. Imagine that instead of receiving movie recommendations from Netflix or holiday recommendations from Expedia, you receive daily job offers from Monster or LinkedIn — and that those jobs are actually right for you.
Now if only we could send our avatars to work while we stay in bed.

Dr Tomas Chamorro-Premuzic is an international authority in personality profiling and psychometric testing. He is a Professor of Business Psychology at University College London (UCL), Visiting Professor at New York University, and has previously taught at the London School of Economics. He is co-founder of metaprofiling.com.