You Should be a Project Manager: Can Neuroscience Help You Find the Right Career?
Take a stroll through AngelList and you’ll see what I mean: Crossfader is “Instagram for music”; Dealflicks is “Priceline for movie tickets”; and my personal favorite — though it only adheres to the spirit of my joke while discarding the strict form — is uBiome’s “big data from bacteria,” because it sounds like a tagline for a biopunk novel.
I suppose this formula caught on because, used correctly, it can be intriguing: put a familiar product in a strange place, and people will want to see what happens.
I’m not immune to it: when Matthew Malter Cohen, head of product at the career-assessment and hiring platform Pymetrics, told me I could think of the platform’s service as “similar to the way Netflix matches movies to your preferences, except we’re matching applicants to the companies’ preferences,” I was hooked. That’s mostly because Netflix’s strangely specific genre suggestions have always amused me (though, they seem slightly tamer in recent days, don’t they?), and I wanted to see if Pymetrics would suggest careers with a similarly fine-toothed approach: “Feel-good nonprofit organizations with a strong female boss,” perhaps?
Unsurprisingly, it turns out that Pymetrics takes its career suggestions more seriously than Netflix takes its movie recommendations, but that doesn’t mean that the platform isn’t a whole lot of strange, bemusing, enlightening fun.
Online Job Hunt 3.0
Currently in beta, Pymetrics uses a series of neuroscience-based games borrowed from research labs to assess your cognitive and emotional profile. It then uses that information to determine which industries and companies are good fits for you. “The way that we do that matching is that we actually have profiles in our database of successful employees in those industries and at those companies, and we match your profile to them,” Cohen explained.
“The idea is that it is available online to any student or jobseeker to use,” Cohen added. “it functions essentially like a marketplace.”
I played the role of intrepid jobseeker, and I can confirm that Pymetrics is every bit as easy to use as Cohen says. You simply pop over to the Pymetrics website, fill out a brief profile or connect your LinkedIn page, and start playing games. The more games you play, the more information Pymetrics has to build a complex cognitive/emotional profile for you. The more complex your profile, the more accurate your career suggestions.
Companies can use Pymetrics as a talent-sourcing avenue. In the same way that the platform gives jobseekers career recommendations, it gives organizations access to a database of candidates who are good fits for them. “And they only see good fits,” Cohen said. “Obviously, you wouldn’t want to be putting information out there that is negative about yourself.”
Let’s say your results suggest that you’d make a great teacher, but a terrible financier. Pymetrics isn’t going to tell the banks you suck; it’s only going to tell the schools they should take a look at you ASAP.
“We just feel like this is a natural progression in terms of online job hunting,” said Cohen. “If you think of the original listings hosted by Monster as job hunting 1.0, and LinkedIn is sort of like 2.0, then Pymetrics is 3.0.”
“You Prefer to Process Things Before Reacting to Them”
(And I suggest you do the same: what follows here is an outline of my own personal experience with Pymetrics. Clearly, one user can’t claim to represent everyone.)
The Pymetrics games are, overall, actually pretty fun, which you may not expect when you hear they’re based in neuroscience. I’d compare most of them to Flappy Bird, but without the frustration : they generally rely on repeated and strategic keystrokes, which make them potently simple and entrancing. But unlike the mindless hypnotism of Flappy Bird, the Pymetrics games are always built around some sort of thought puzzle, so there’s a good deal of cognition required on your part.
It’s also not always immediately obvious how the games are going to provide insight into your person — e.g., Pymetrics was clearly inspecting my emotional intelligence when it asked me to determine the emotions people were feeling based on pictures of their eyes, but the game in which I pressed certain arrow keys depending on the colors and directions of little arrows that popped up onscreen? Still not sure about that one, but I don’t necessarily think that’s a problem: trusting your results is a matter of trusting Pymetrics’s scientific process in total, not a matter of knowing what each step along the way means. Throwing out your Pymetrics profile because you don’t understand a couple of the games you played is like discrediting the results of the Stanford marshmallow experiment because the children who participated didn’t understand why they were being asked to not eat marshmallows.
But I’ll admit: whether or not you can readily place your trust in Pymetrics might be a matter of personal conditioning. For those who understand and/or enjoy scientific objectivity, it will be pretty easy to accept Pymetrics’s revelations precisely because they are relatively untainted by human interpretation. For others, the idea that games will tell you things about yourself —and possibly things you didn’t know, things you may not agree with — is worrisome. “I know what I like to do and what I’m good at,” the thinking goes, “why should I listen to a website?” I feel like a lot of people will fall somewhere between the two poles.
It’s a little silly that I can even anticipate this sort of conflicted thinking about Pymetrics, but we’ve seen how easily the general population can misunderstand science. At some time or another, most of us have been one of those baffled, terrified people. I for one still have no idea who I’m supposed to listen to about GMOs. But my point is this: anytime science tells us something we don’t want to hear, there will be people who react negatively. That doesn’t mean Pymetrics should give up — no, that’s all the more reason for Pymetrics to keep fighting the good fight. To quote Neil deGrasse Tyson, “That’s the good thing about science: It’s true whether or not you believe in it.”
Now, about that section title: upon completing any game, Pymetrics offers you what they call an “insight”: a direct, one-sentence summary of what the platform learned about you from that particular game.
While the insights themselves aren’t always illuminating (besides, the real meat is in the reports), they are fun to collect. In fact, they were the number one reason I spent close to two hours “researching”: I took screenshots of all of my insights, and I now proudly have a whole stash in a folder on my desktop.
Some of the insights I received were things I already knew about myself, like the one I used to title this section. But I’d say an equal number of the insights were surprising — sometimes, greatly so.
What Pymetrics Knows
If we take the technology metaphor (“online job hunt 3.0”) and run with it, we can say that, by utilizing neuroscience, Pymetrics hopes to fix a few bugs that currently exist in sourcing, hiring, and job-seeking processes. In fact, the company came about precisely because cofounders Frida Polli (CEO) and Julie Yu (CSO) were unhappy with the methods they saw people using to determine their career paths.
According to Cohen, Polli and Yu, both accomplished scientists, met during their postdoctoral days. Polli decided to leave academia, because she wanted to start a business. She went to Harvard Business School (HBS) for her MBA. While at HBS, she took a few of the traditional, questionnaire-style career assessments: the MBTI, the Strong Interest Inventory, the Holland Code Career Test, etc.
“What struck [Polli] was that [the tests] are all question and answer,” Cohen said. “They’re all asking you essentially about your interests and motivations to tell you what you’re good at. From a decade in the lab, being trained as a neuropsychologist, it was very clear to her that this isn’t necessarily the best way to understand what someone’s traits are or what someone is good at.”
Consider this bug No. 1: questionnaires can help you aggregate and solidify the things you already know about yourself, but they can’t reveal anything new about you, because “the only input into the questionnaire is what you tell it,” to quote Cohen.
The problem with relying on our self-knowledge is that it’s usually inaccurate, so how can we trust that we know ourselves well enough to figure out which career really, truly suits us best? It’s not like no one’s ever landed a job that was totally wrong for them before.
So while Pymetrics did tell me some things I already knew about myself, it also uncovered some character traits I had never picked up on. After I played a game in which I had to repeat sequences of random numbers, Pymetrics told me, “If you don’t write them down, the details elude you …” Ominous ellipsis aside, I loved this insight because it was something that I never realized about myself, but I could still immediately see that it was true. My need to record all my interviews so that I can’t possibly misquote anyone; the notebooks I filled in every course I took at college; the endless to-do lists I type up on my PC: it all adds up — I don’t like keeping details in my head.
And then there was the time that Pymetrics got mad at me (or, at least, seemed to): I played a game in which I had to divide money between myself and a stranger. I used the same strategy every time: make it so that everyone ends up with an even amount of money. Even when given the option to take money from people, I kept everything even. At the end of it all, Pymetrics subtly scolded me the way a sitcom-stereotype spouse might chide her husband for never asking his boss about the raise he deserves: “Using a situation to pursue your interests can be a real strength.”
What you Definitely Don’t Know
Bug No. 2, according to Cohen, is that questionnaires can be faked. “Just like with resumés, you can sort of understand, if you’re an intelligent person, what the person or assessment is asking you to get at. So you can tell it something that gets you the result you desire,” Cohen said.
As an example, Cohen offered a hypothetical jobseeker looking for a role in finance. In the financial world, it’s desirable to have employees who respond to and enjoy monetary rewards. By strategically answering certain questions, the jobseeker can game the questionnaire into returning a result that says he likes monetary rewards, even if that isn’t true. Now the company is getting an employee who isn’t actually a great fit, and the jobseeker is getting a job that they’ll probably end up loathing.
Earlier, I mentioned playing a game that asked me to guess the emotions people were feeling based on pictures of their eyes. After completing the game, Pymetrics gave me some unsettling news: “You don’t let other people’s emotions affect you.” Ouch. Am I a psychopath? Probably not, but I always thought I was super empathetic.
But that’s part of Pymetrics’s mission, isn’t it? To uncover what we don’t know, and to correct what we think we know. Had this been a questionnaire, I would have played up empathy in my answers — I would have subconsciously gamed the system, based on a faulty perception of myself. But I didn’t have that chance, and it turns out I may need to adjust my self-image accordingly, or maybe focus on building up the empathy I lack but want. Either way, you don’t expect a career assessment platform to knock you out of your orbit, but Pymetrics set me on the path of some serious soul-searching with this result.
It’s results like this that will, I think, lead to some of that conflicted thinking that I mentioned above: my gut reaction was to dismiss this result as a mistake. It didn’t square with my beliefs, so I wasn’t going to listen to it. Sure, even neuroscience-based games can make mistakes, but maybe this wasn’t one? I can either approach it as a learning opportunity (regardless of how empathetic one is, it’s almost always a good idea to try to be even more empathetic), or I can throw out the whole platform because it said something I didn’t like. Personally, I think users who take the first approach tap into more of Pymetrics’s value.
Note: I don’t think this sort of conflict is Pymetrics’s fault. As I said earlier, I think it’s simply something that happens when scientific principles enter our personal lives: we’re not used to having them there, and it’s up to us to figure out how to adjust.
Bug No. 3: unconscious biases are a real problem in the world of hiring, and they’re especially hard to overcome because they’re so hard to see. By allowing neuroscience to evaluate candidates’ fits, Cohen hopes Pymetrics can prevent unconscious biases from ever entering the equation: “We fit applicants to an industry or company based on their behaviors in the game set. We do not look at things like what your name is, what your gender is, how old you are,” Cohen said. “All of these characteristics can often unfairly bias a recruiter’s or hiring manager’s opinion of a resumé outside of even their awareness.”
I can’t personally speak to this, as I’m a white male, and therefore not the target of such bias. But I can see how Pymetrics could circumvent these biases, and that’s something I’m really excited about.
Where to Now?
The first thing I noticed about Pymetrics’s results reports is that they’re gorgeously designed. The cognitive/emotional information is organized along colorful, manipulable nodes. The “industry fit” section offers a circular chart that breaks down how well you fit into a multitude of possible careers. Aside from these visual presentations, both reports also offer detailed verbal analyses of your results.
The results are easily the most impressive thing about Pymetrics. I was especially fascinated with my cognitive/emotional profile, which gave me percentile scores for things like “trust” and “attention” — things that we often think of as thoroughly unquantifiable. For example, I learned that I’m exceedingly trusting, with a 100 percent score in that category. Some might say that makes me naive, but I wear it as a badge of honor.
These wider categories are broken down to offer more specific insights. For example, I wasn’t merely told I was altruistic; I was told that I’m altruistic after stressful situations. Similarly, my risk/reward results were enlightening in their thoroughness: I now know my risk preferences for high-, medium-, and low-risk situations, as well as how I learn from different types of risks and how well I deal with ambiguity. And though the cognitive/emotional profile Pymetrics built for me contained a lot of new information, people close to me confirmed the accuracy of my results.
Of course, this is a careers-oriented website, so what we really want to know is: how are the career suggestions?
Like my personal results, I found my “industry fit” section full of surprises. My top three matches, in order, were: project management, technology, and education. About that third result: I tried my hand at education for a couple of years, and it was not my thing at all. Does that invalidate the rest of Pymetrics’s suggestions?
I’d say no. I checked out the profile of successful educators on Pymetrics, I have all of the described traits. I can also say that every great teacher I worked with had these traits, too. So the educator profile is accurate, and my personal results are accurate — but life is weird. Things don’t always work out the way they should. Pymetrics’s neuroscientific approach isn’t supposed to be infallible; when it comes to people, we can never totally avoid error.
And what about my other career recommendations? I’ve never considered project management or technology before, but now I’m interested. Ultimately, this is the most significant advantage Pymetrics offers over other career assessments: it uncovers things you wouldn’t otherwise learn about yourself. It’s not the god of career placement (nor does it claim to be), but it’s a particularly keen observer.