This post originally appeared at LinkedIn. Follow the author here.
I know myself, and employers know what they want to hire, but how do we explain that to each other efficiently and accurately? The marketplace for people and jobs is broken, especially for the small businesses that create the bulk of jobs in the United States. And it’s part of why so many people are out of work while simultaneously so many jobs are unfilled. Unemployment is an information asymmetry problem.
And it’s the one thing I’d fix if I could.
Information asymmetry is when one party has better information than another party. Let’s say I’m selling my car and I know the passenger door rattles when I drive over 65 mph. You are buying my car, and have no idea. That’s information asymmetry.
Job-matching efforts also suffer from information asymmetry, or what I call the Color Blue Problem. How do I know that when I see the color blue, it’s the same as when you see it? How do I know that when I describe myself to an employer, they know what I mean? And that when a hiring manager describes what she wants in a job posting, how do I know what she means?
In practice, it looks* like this:
Resumes stink. They’re a simply awful way of marketing yourself for a job. Some of that is our fault as job-seekers and can be fixed, as I wrote here and here. But an employer has no way of knowing if most companies on a resume are good or bad (is working at “LaszloCo” a good sign?), if a title means anything (VP is a senior title in tech, but not in banking, and even in tech some companies have one VP for every 20 people and some have one per 300), or even what my words mean (is a “superb programmer” the co-inventor of Google or just really, really good at Logo?). And employers are completely blind to the indefinable things that make you “you,” such as generosity, curiosity, or playfulness.
It’s just as bad on the job-posting side. Job descriptions are often written from generic templates, don’t give you a sense of what the job truly requires or what would make you successful in it, and are just plain boring. Here’s an insider’s view of what the process feels like from the other side:
Resume screeners and interviewers deliver the coup de grace: We all think we are great at assessing candidates. We’re not. We are biased, ask bad interview questions, rarely go back and check if our predictions were correct, and so on. We only hire the best, right? Then how did all those slackers in Sam’s department get hired? More to come on this in a future post, but the point is that the job-matching process is fragile and error-prone.
The root cause is that we can’t convey perfect information about our own skills, nor can employers convey perfect information about what they need. We both say the job is a “Color Blue” job, but we have no way of knowing for sure if we both mean the same thing when we say “Color Blue.” Information asymmetry.
The enormous opportunity to solve unemployment
But what if you could perfectly convey the real you? Not just your training and feats, but in what kind of workplace you would thrive. Whether you like to work alone or in groups. Whether you are a specialist or a utility player. Exactly how good you are at your disciplines. And what if sending this message was believable? If a prospective employer could know with certainty that they can see the real you.
Now, what if you had the same insight into jobs? Is my prospective manager a control freak or checked out? Is this job a stretch, just right, or completely out of reach? Do I have the general attributes that will set me up for rapid promotion, or will I be stuck in the same job for a decade? Do givers or takers thrive in this company?
In the short-term, much unemployment could be eliminated by doing a better job of matching people and jobs. By solving the Color Blue Problem.
There has been a visible revolution in the ability to analyze lots of data. Less noticed are advances in organizational science and behavioral economics, ranging from Amy Wrzesniewski’s pioneering jobcrafting work, to Evolv’s work on matching people to jobs, to Googler Brian Welle’s work on unconscious bias. (Disclosure: I was until recently a board member of Evolv and of course work at Google.)
Mapping the reality of what you have to offer against the reality of what organizations need—and who will thrive in that specific context—is a hard problem. But it is solvable. It becomes possible to move beyond “four years of public accounting experience” to “ability to learn quantitative methods combined with a zeal for catching and correcting the smallest of errors, persuade with data, and thrive in social settings” as job criteria, and to then identify people based on who they really are. For individuals, it becomes possible to find roles where they will excel regardless of where they went to college, or even if they went to college.
Now, imagine this works. If you’re a welder in Detroit, you can find out what skills are increasingly or decreasingly in demand. Then you can make some informed choices: Should I move to Atlanta where there will be more welding jobs, or stay put and go to nursing school since I know there will be demand for those jobs at home? If I go back to school, which schools’ graduates are most likely to end up in the jobs that I want?
Slowly we’d become able to not just match people today, but also to tell people where to invest to be ready for tomorrow’s jobs.
Hundreds of billions of dollars are spent each year on recruiting, so there’s a lot of incentive to figure this out. The trick is you can’t do this by conducting exhaustive (and exhausting surveys) coupled with anthropological dissections of every group inside every organization. Not practical.
The most efficient way is by looking at large sets of data and inferring relationships, similarities, and predictors of success and failure. And the only way to do that is with permission, appropriate privacy safeguards, and enough value delivered to the individuals and organizations to make them want to take part.
From a business perspective, the promise of solving unemployment is enormous. From a social perspective, it’s exhilarating. And from a computer and organizational science perspective, it’s coming into reach.