Lunch With Coworkers: The Forgotten Performance Multiplier

The single best-documented predictor of how well a team performs is not what tools they use, what their org chart looks like, or what their compensation structure is. It is whether they eat together.

This sounds like a soft claim. It is not. The research is robust, the effect sizes are large, and the mechanisms — trust, social capital, informal information flow — are exactly what makes teams work.

Here is what the research actually shows, and what it means for how you spend your lunch hour.

The Firefighter Study

Citation: Kevin Kniffin, Brian Wansink, Carol Devine, Mitsuru Shimizu, "Eating Together at the Firehouse: How Workplace Commensality Relates to the Performance of Firefighters," Human Performance, 28(4), 281–306 (2015).

URL: https://doi.org/10.1080/08959285.2015.1021049

Kniffin and colleagues (then at Cornell) surveyed firefighters across 50 US fire stations on their workplace eating practices. They then independently rated each station's performance based on supervisor evaluations.

The finding: Stations where firefighters cooked and ate meals together performed significantly better than stations where firefighters did not.

The effect held after controlling for station size, regional differences, demographics, and tenure. The mechanism the researchers proposed: shared meals build the trust and informal communication that high-performing teams need.

In their own words:

"Sharing meals is a particularly powerful form of commensality that promotes cooperation and team effectiveness."

The implications generalize beyond firefighting. The mechanism — shared meals build trust that translates to team performance — applies to any team where coordination matters.

The Communal Eating Study

Citation: Robin I.M. Dunbar et al., "Breaking Bread: The Functions of Social Eating," Adaptive Human Behavior and Physiology, 3(3), 198–211 (2017).

URL: https://doi.org/10.1007/s40750-017-0061-4

Dunbar (Oxford) is best known for "Dunbar's number" — the 150-person cognitive limit on stable social relationships. This paper applied a similar lens to shared meals.

The finding: People who eat with others more often report higher life satisfaction, larger social networks, greater trust in the people around them, and stronger feelings of community.

Specifically, the study found:

"The more often people eat with others, the more likely they are to feel happy and satisfied with their lives."

The effect was robust across age groups, income levels, and cultural backgrounds. The frequency of shared meals — not the quality of the food, the venue, or the duration — was the active variable.

For workplace contexts, the implication is direct: a team that eats together 2–3 times per week is running a different long-term experiment than a team that does not.

The Anthropological Foundation

The Kniffin and Dunbar findings are the modern empirical confirmation of an older anthropological tradition.

Mary Douglas, the British anthropologist, argued in her 1975 essay "Deciphering a Meal" that shared meals are:

"A microcosm of social order, where classifications of food and timing express and reproduce social relationships."

Translated into modern workplace language: how a team eats together (who shows up, who sits next to whom, what gets discussed) is information about how the team works together. The meal is data.

Claude Fischler, the French sociologist, captured the same idea in L'Homnivore (1990):

"To share food is to share trust. Commensality builds and signals belonging to a group."

Carole Counihan, the cultural anthropologist who spent decades in Florence documenting pranzo:

"Through pranzo, families created and recreated family every day, reinforcing bonds and transmitting values."

The pattern is universal. Across cultures and across centuries, the shared meal is a primary site of relationship-building. Workplaces that capture even a small version of this produce better outcomes than workplaces that don't.

Why It Works

The mechanisms are not mysterious. Five things happen when teams eat together that don't happen otherwise.

1. Informal information flow

Most of what a team needs to coordinate is not in the official meetings. It is in the off-the-record details — what's actually blocking someone, what's not getting attention from leadership, what the customer really said. This information flows through informal channels. Shared meals are one of the best of those channels.

2. Reciprocity

Sharing food activates ancient reciprocity instincts. People who eat together feel a low-grade obligation to each other that translates to easier help-seeking and help-giving on work tasks. The effect is mild per meal and substantial over months.

3. Identity

The team becomes a unit, not a collection of individuals. The "we" framing strengthens. This shows up in how people talk about the team to outsiders and in how willing they are to take on group-level commitments.

4. Conflict de-escalation

Disagreements over shared meals are different from disagreements in conference rooms. The food, the table, the casual setting all dampen the intensity. Teams that eat together regularly fight less because they have a non-conference-room context in which to discuss difficult things.

5. Recovery transfer

Even when lunch conversation is partially work-related, the social context provides Sonnentag-style detachment from individual task focus. The cognitive load shifts from "solving the problem on my screen" to "engaging with this person across the table." That shift is itself restorative.

The Counter-Intuition

The research suggests something many workplace cultures get wrong: communal lunch is not optional or supplementary. It is performance-relevant.

The companies that have figured this out — partly Pixar (with its central atrium designed to force casual interaction), partly Google (with its food-on-site culture), partly any well-run small company — treat the shared meal as infrastructure. They invest in it the way they invest in conference rooms or office furniture.

The companies that haven't figured this out treat lunch as a personal matter. Workers eat at their desks alone. Managers schedule "lunch and learns" that aren't really lunch. The shared-meal infrastructure exists in the org chart only as a coffee machine.

The performance cost of the second model is real and accumulates over time. Teams without commensality become collections of individuals who happen to share a budget. Teams with commensality become teams.

How to Borrow This Without Being That Person

The practical implementation has to navigate two cultural risks.

Risk 1: forced fun

The team lunch that gets scheduled because the manager read a productivity article is worse than no team lunch. Forced communal eating is low on the "control" dimension of the Sonnentag framework — it produces less recovery than autonomous eating and can actively damage relationships.

Risk 2: cliques

Repeated meals with the same small subgroup can intensify in-group / out-group dynamics. A team where 4 people always eat together and 6 always eat separately is worse than a team where everyone eats separately.

The practical path

Three rules that produce most of the benefit without the failure modes:

1. One or two communal lunches per week, optional. Not five. Not zero. Two is the threshold the Dunbar data suggests is sufficient for the social-bonding effect.
2. Rotating composition. Not always the same people. New combinations build new ties.
3. No agenda. A "lunch and learn" is a meeting with food. A lunch is not a meeting. Treat them differently.

The hardest one is the third. American workplace culture has a strong impulse to make every shared time productive. The commensality research says this is exactly backwards: the productivity comes from the meal not being productive, not from making it productive.

What This Means for Your Calendar

For the individual worker, the operational implication: don't default to solo lunch every day. Build in 1–2 communal lunches per week as a habit, and protect the other 3–4 as solo time.

The calendar implication: both kinds of lunch need defending against incoming asks. The communal lunch on Wednesday is just as easily overridden by a 12:30 meeting as the solo lunch on Friday. Both need camouflage.

This is why CovertLunch is the same product for both use cases. The calendar event looks like a real meeting in either case. What you do during the protected hour is up to you.

Try CovertLunch free for 7 days →

A Cross-Cultural Note

The American workplace is structurally hostile to communal lunch. Open-plan offices, dispersed remote teams, calendar-driven schedules — all push toward solo meals at desks.

The European workplace, particularly in France, Italy, and Spain, is structurally supportive of it. The cultural defaults work for you, not against you.

In countries with the structural support, the commensality research is less surprising because the practice is already in place. In the US, applying the research requires conscious effort and protected time. The protected time is what we've been arguing for throughout this content set.

The hour is not just yours. Some of it can be shared. The sharing is the part that the research keeps finding matters most.

Related Reading

• What Japan's bento culture teaches us about office lunch
• Why French workers get an hour for lunch
• The Lunch Protection Playbook

Frequently Asked Questions

What is commensality?

The practice of eating meals together. From the Latin com- (with) and mensa (table). Cultural anthropologists use the term to describe meal-sharing as a social institution distinct from eating itself.

Does eating with coworkers really help team performance?

The Kniffin 2015 firefighter study is the cleanest evidence. Stations where firefighters cooked and ate together performed better on supervisor-rated performance, controlling for size, region, and demographics. Effect size was substantial.

How many communal lunches per week?

The Dunbar research suggests the bonding effect emerges with regular shared meals (not necessarily daily). For most workplaces, 1–2 communal lunches per week is realistic and produces most of the benefit.

Should I make team lunches mandatory?

No. Per the Sonnentag autonomy research, mandatory lunches produce less recovery than chosen ones. Optional, with cultural encouragement, is better than required.

What about remote teams?

Remote teams face structural disadvantages on commensality. Possible substitutes: video-lunch sessions (1x per week, optional), in-person quarterly gatherings, asynchronous food-sharing rituals (recipes, photos). None are as effective as in-person shared meals, but they are better than nothing.