Science Says Your Friend Group Does Not Need Specialists

Everyone has a mental roster. The friend you text when a date goes sideways, the one you call for a pep talk before a job interview, the group chat that exists purely to make you laugh. Somewhere along the way, wellness culture turned that roster into advice for your friend group: curate your circle, it says — assign each person an emotional specialty, and your friendships will run like a well-staffed team.

A new study put that intuition to a serious test, and the results are not what the self-help shelves would predict. Researchers at the University of Würzburg ran what they call the FRIENDS study, tracking real friendships in real time to see whether your friend group genuinely assigns specific friends to specific emotional needs — and whether that matchmaking pays off. According to the research team, whose findings appeared in BMC Psychology, the specialization people believe in turned out to be faint, lopsided, and — most importantly — pointless for actually feeling better.

The method is what makes the finding hard to dismiss. In a preregistered 14-day ecological momentary assessment, 106 adults received four semi-random phone prompts a day asking about a recent contact with a friend. Each of the resulting 2,796 reports captured which friend was involved, which of six motives — problem solving, encouragement, reappraisal, humor, distraction, or touch — the participant had been seeking, and how the interaction actually went.

The specialization barely showed up

On the surface, the data looked like the toolbox theory was right: people did seem to route different needs toward different friends. But the analysts separated mere frequency from true specificity — because a best friend you contact for everything will look "specialized" for every motive by default. Once overall involvement was stripped out, and the real patterns were tested against thousands of randomized reshufflings of the same contacts, the picture collapsed. Only two motives — practical advice and distraction — showed genuine channeling toward particular friends. For the rest, the apparent pairing could have been chance.

Zooming out to whole friendship networks told the same deflated story. A network-level specialization index did beat chance on average, but the effect was modest — the authors report a standardized effect size of d = 0.36 — and it concealed wild individual variation: only 18 of 104 participants had networks whose specialization reached statistical significance. As the authors put it, specialization seems to be a property of some people's networks rather than a universal architecture of friendship.

And where it existed, it bought nothing

Here is the part that should change how you think about your group chat. Neither the partner-motive pairing nor the network-level indices had any measurable relationship with the outcomes that actually matter: how effective people found the contact, how positively they rated the interaction, their psychological well-being, or their loneliness. That null result held across every alternative way the team sliced the data, according to the authors. Wherever the toolbox pattern appeared, it bought its owners nothing.

The authors draw a pointed lesson from that emptiness. "The findings caution against inferring functional specialization from partner-selection data alone," they write — a warning aimed squarely at the way we all read our own social lives. Much of what passes for specialization, the analysis suggests, is just concentration: a few favored people absorbing most of the contact.

What the study could not see

To be fair, the researchers flag a real blind spot of their own. The design captured what participants sought and experienced — but did not record what the friend actually did during the exchange. The comfort that landed might have come from a completely different strategy than the one the person reached out hoping for. Future work, the authors say, will need to observe the partner's side of the interaction to understand when turning to friends genuinely helps. Coverage of the paper in Scienmag highlights the same caveat.

That caveat matters, but it does not rescue the specialists theory. The study's design was conservative by choice — preregistered hypotheses, chance baselines built per person, results robust to every alternative specification — which makes the null finding harder to wave away than a single failed survey.

What to do with your friendships instead

So if you do not need a friend for every mood, what does the evidence actually support? Something refreshingly low-maintenance. The finding lines up with a broader thread in relationship science: the value of a social network comes less from optimal matching than from having people to turn to at all, rather than routing the right emotional job to the right person.

That is good news for anyone whose friendships feel messy and unoptimized. You do not need to audit your contacts, rank their utility, or quietly demote the friend who gives mediocre advice but shows up every time. The friend who shows up every time is the whole point. And when noise gets in the way of those connections, small tweaks help: recent reporting found that a small social media cut can strengthen friendships. Keep the group chat, call whoever picks up, and stop treating your friend group like an org chart — the data suggests the org chart was not doing the work.