It is widely accepted that team diversity is good for performance — that bringing a variety of perspectives and problem-solving approaches leads to greater creativity and better outcomes. The intuition is right. But the supporting research is more nuanced than most discussions of "diversity" admit, and the nuance matters when you are designing teams.
A substantial body of research has actually found that team diversity can create social divisions, which produce negative performance outcomes for the group. So which is it?
The measurement problem#
The conflicting findings have a common cause: most studies — and most management conversations — define diversity primarily through demographic variables like race, age, and gender. Those variables matter for many reasons, but they are not what is doing the predictive work when diverse teams outperform homogenous ones.
A growing body of research, including a fascinating study by Greg Kress, then a PhD student at Stanford, points in a clearer direction: team members benefit not from demographic variation per se but from variation in work styles — the ways they think, solve problems, and communicate.
What stood out in the Stanford study#
Kress studied fourteen work-style variables across Stanford masters-level project teams of three to nine people. One variable showed a particularly strong correlation with project performance: Extraverted Feeling, a cognitive mode in the Wilde-Type Teamology framework.
Briefly: a person has a preference for extraversion if they get energy from working with others; they have a preference for feeling if they interpret experiences emotionally rather than purely rationally. People high on Extraverted Feeling tend to be quick to sense and address conflict, and they tend to work actively to keep communication lines open within the team.
That second behavior is the one that matters. Teams that included an Extraverted Feeling member outperformed teams that didn't, holding other variables constant, because someone was actively maintaining the social infrastructure that lets disagreement be productive instead of destructive.
The practical takeaway#
Two things flow from this for anyone designing teams:
1. Diversity is about cognitive and behavioral range, not just demographic composition.
The teams that win the upside of difference are the ones whose members approach problems differently, communicate differently, and care about different things. Demographic diversity often correlates with this but doesn't cause it. Building a team that looks diverse but thinks identically is a known failure mode.
2. Communication maintenance is not optional.
A diverse team without someone who actively keeps the communication lines open is the version of diversity that produces social divisions and lower performance. You can either select for that role explicitly — looking for people high on the equivalent of Extraverted Feeling — or you can institutionalize the behavior through team rituals. Ideally both.
How this shows up in Zeswa#
Our compatibility model treats work-style range as a first-class input alongside skills and values. When you bring a new candidate into an existing team, we surface where they extend the team's cognitive range, where they overlap, and where they fill a gap in the team's communication maintenance.
The goal is not to maximize similarity. It is also not to maximize difference for its own sake. It is to build the configuration of attributes that the evidence says actually predicts performance — and that is almost always more interesting than the conversations we are still having about diversity in the abstract.
Research conducted at Stanford University on first-year masters-level student teams of three to nine members. Findings should be read alongside the broader team-composition literature; effect sizes vary by task type and team interdependence.