Vitaliy Katsenelson, Chief Investment Officer at Investment Management Associates (Denver, Colorado) recently tweeted that they have added, in their job interview process, to check the candidate’s Uber rating (shared by the candidate, on-demand/request, by IMA)!
Reason? “If you treat strangers with respect then you will treat customers and coworkers well too.”
I was immediately reminded of two related instances.
First, a scene from Lage Raho Munna Bhai, where Sanjay Dutt’s Munna Bhai is taking calls from listeners and is helping them solve problems. A girl calls him and says that her dad has chosen a prospective groom for her through the matrimonial section of a newspaper and she is to meet him at a restaurant. And she wonders aloud how she can decide anything about the guy with just one meeting. Remember what Munna Bhai asks her to do (suggested by Gandhi Ji, of course)? Take a look 🙂
The second instance is an episode from Black Mirror… the first episode in season 3, called Nosedive. The episode is set in a world where people are supposed to rate each other on a scale of one to five, for every single social interaction! This is an exaggerated version of something very real already – credit scores. But while credit scores are generated and sold by few private companies (in India, at least) based on our financial transactions, the Black Mirror episode took it to an absurd extreme by combining social media rating, feedback, and things like the Uber score.
Vitaliy’s idea to include the Uber rating of a customer is equally misleading as a metric to use during the hiring process. It may sound great in a movie when seen in action as a one-on-one reaction like that in Munna Bhai, but in real life, it falters on multiple levels. Vitaliy found this out too, as he explained the follow-up in a Financial Times piece.
Almost all of our digitally recorded social interactions, whether in an Uber cab that reflects in the app, or via any of the social media platforms, are poor signals of assessing individuals for employment. Or, in other words, the kind of assumptions you may make through those signals may be misleading within the context of recruitment.
There are many reasons for this.
1. A performance
The numbers, and social media utterances, do not showcase the larger context that you could decipher only by observing the person a lot more, in closer quarters and that too, without them knowing that you are watching them. When they are aware that people are watching, and judging (as in any social media platform), their behavior could be akin to how we behave when we are on the stage.
When we are on the stage, we perform, no matter what we are on the stage for. Are we acting, when performing? Not really. It is still us, but the difference is discernible – like your talking to a neighbor at a company meeting vs. you addressing all the meeting participants. There is a marked difference between both versions of you.
Social media utterances are hence a better version of your real, normal self – a projected self. You tend to play to the gallery, literally.
2. A snapshot of the real self
This doesn’t mean people are not themselves on social media either – a lot of people share the most intimate things about themselves on social media without even considering how total strangers would perceive those words. But then, these could simply offer a fleeting glance about the person that you may use to form very, very rudimentary opinions about the person, definitely not use them to decide whether they can be hired or not!
3. Recruiter bias
Besides all this, there are several other problems. For instance, the recruiter’s predisposition itself is a problem. Assume that a recruiter likes cats and she finds a candidate talk profusely about cats. That could color her opinions too, even if that has nothing to do with the skills needed for the job!
4. Tools and algorithms
Things get considerably worse when recruiting teams use tools and algorithms that assign scores as they trawl through a candidate’s online footprint, most of the time without the explicit consent of the candidate (this is illegal in many countries, though such illegality may be laughed at, in India). The algorithms are likely to miss context and use keyword-related checks to make broad assumptions about a candidate and approve/disapprove them for the next level.
A better way to use such algorithms may be to not just form a broad character map through the online footprint but to tweak them to deduce the appropriateness for the job being applied for. That’s a lot more difficult than just a keyword-based check, but more useful too, if achieved.
About that Uber rating
With an Uber rating, things are even more muddled since there is no context of why a/many drivers rated you the way they did. There is so much context missing in Uber trying to distill all that into a single number from 1 to 5! Take, for example, women who may be disproportionately affected by rating a driver poorly because they did not feel safe during the cab ride (which is very, very real). See this Twitter thread, for context:
Tons of women responded too, on similar lines, to Vitaliy’s original tweet.
Beyond all this, there is research that showcases that social media research about a candidate, regardless of what the recruiters think or infer from it, “appears to be unrelated to future job performance or withdrawal intentions”! That’s reason enough to reconsider the inferences from a candidate’s social media profile.
The new research from Singapore Management University titled, ‘What’s on job seekers’ social media sites? A content analysis and effects of structure on recruiter judgments and predictive validity‘ suggests this.
A related point is what is called the dilution effect, a judgment bias in which people underutilize diagnostic information when nondiagnostic information is also present. The presence of social media signals does not really help assess if a person would perform their jobs well, but they inundate the recruiter with so many other kinds of information that tend to make the recruiter feel that they are conveying something (even if they do not). Much of the reasons behind checking social media posts are less about on-job performance, or about the potential of a candidate to stay in the job, and more about team participation and social cohesion.
However, there is one area where social media footprint checks could indeed help – decipher outright problems. For instance, visible/vocal association with violent activities or groups! One, it would be stupid for the candidates to flaunt such association in the public, online, but even here, there is a grey area – when a candidate has participated in anti-Government activities offline, with or without incitement to violence, and then the Government changes (to the one the candidate is supporting)… how would that be seen? 🙂
It’s one thing to decide to follow or unfollow someone on social media platforms based on your assumption of who they may be after you go through their social media feeds… and it is entirely something else to use those assumptions to propel hiring decisions! The latter does happen (I know of a lot of recruiters who use this method to filter candidates), but it may be useful to remember that it is misleading.