Most people assume that getting ahead at work comes down to doing the work well. Solve the hard problem, hit the number, ship the project, and the evaluation should follow the results. A 2021 study published in PLOS ONE by Lucie Nikoleizig, Stefan Schmukle, Maurin Griebenow, and Sascha Krause tested that assumption directly, and the results were more complicated than a simple story about merit. The researchers put 164 participants into 41 small groups, had each group work through a problem-solving task together, and then had both the group members and trained external observers rate how well each person had performed. Task competence did predict how someone was rated, but it wasn’t the strongest predictor. Two other factors, how much someone spoke during the group task and how physically expressive they were while speaking, explained even more of the variation in performance ratings than actual competence did. Likability added a further, independent layer on top of all of that. In practice, two people who solved the problem equally well could end up with meaningfully different performance evaluations depending on how much they talked, how animated they were, and how much the group simply liked them.
That finding lines up with an older and more widely cited piece of research on how likability functions in professional relationships more broadly. In 2005, Tiziana Casciaro and Miguel Sousa Lobo published a study in Harvard Business Review examining roughly 10,000 work relationships across five organizations, sorting coworkers into four rough types: the competent jerk, who gets results but is unpleasant to deal with; the lovable fool, who is well liked but not especially skilled; the lovable star, who is both; and the incompetent jerk, who is neither. Across every organization studied, people consistently preferred working with a merely likable colleague over a more capable but difficult one. Given a real choice about who to bring onto a project, competence alone wasn’t enough to overcome unpleasantness, and likability alone was often enough to make up for a shortfall in skill.
There’s a reasonable question about why this would be true, and the honest answer leans on a mix of established psychology and plain practicality rather than any single decisive study. Working with someone is not a one-time transaction. It involves repeated interactions, shared stress, and a fair amount of interpretation, since most tasks involve some ambiguity about what “good work” even looked like in the moment. A well-liked colleague gets the benefit of the doubt more often when something is unclear or goes wrong, while a disliked one is scrutinized more closely for the same mistake. None of this means likability is some kind of trick that overrides reality entirely. The Casciaro and Lobo research also found that nobody, in any organization studied, preferred working with an incompetent jerk. Being disliked and unskilled at once was a bad combination in every setting. Likability functions more as a thumb on the scale than a replacement for doing the work.
It’s worth being cautious about how far to take findings like these. The Nikoleizig study, while methodologically careful, is a single piece of research built on a specific set of laboratory groups doing a specific kind of task, and it can’t tell us exactly how much any of this generalizes to, say, a years-long career at a large company where formal metrics and paper trails carry more weight than they do in a short group exercise. The Casciaro and Lobo data, though drawn from real organizations, is now two decades old and predates a lot of change in how workplaces evaluate people, including a heavier reliance on structured performance reviews designed partly to reduce the influence of exactly this kind of bias. Both studies describe a real pattern, not an iron law, and neither claims that likability outweighs competence in every setting or every kind of job.
A separate, more macro-level thread of research adds useful context about why social skills matter at the level of the whole labor market, not just individual evaluations. A Pew Research Center analysis of nearly four decades of U.S. Department of Labor data found that employment has grown fastest in occupations that place a premium on skills like negotiation, persuasion, and active listening, even as automation reshaped which technical skills are in demand. Somewhat counterintuitively, jobs requiring the most repetitive analytical or mechanical work have not been where employment growth concentrated. The jobs Americans have moved into over time have disproportionately been ones where dealing well with other people is part of the core job description, not an incidental extra.
None of this is an argument for performing warmth instead of doing good work, and none of the researchers behind these studies frame it that way. What it does suggest is that social skills aren’t a soft add-on to a career built on competence. They appear to function as part of how competence actually gets recognized, communicated, and rewarded in a workplace built around other people rather than isolated, objectively measured tasks. Being good at the job and being someone people want to work with turn out to be more entangled than the tidy version of meritocracy usually assumes.