Work Friendships: What the Research Says About Having Friends at the Office

Friends at work can lift your performance and wear you out, often both at once. In a 2016 study of insurance employees published in Personnel Psychology, more work friends went with better ratings and more exhaustion, and a second sample of retail and restaurant staff tied the upkeep of those friendships to weaker performance.¹

Work friendships are a mixed blessing, and on balance the research leans in their favor. They are linked to engagement, job satisfaction and staying in a job, though nearly all of that evidence is correlational, so it cannot show whether friendship makes a job better or a good job breeds friendships. The costs include emotional strain, clashes between friendship and job roles, and hurt when a friend is promoted or leaves. The habits that limit those costs are small and draw on the same skills as the rest of emotional intelligence at work: agree on how much personal talk fits the workday, talk about possible rivalry before a promotion decision, keep decisions out of private chats, and, if you manage a friend, write your criteria down before you decide.

What work friendships are linked to, and what those links can’t prove

Work friendships are linked to higher employee engagementemployee engagement: A term with two main meanings: in research, a positive state of energy, dedication and absorption in one's work; in workplace surveys such as Gallup's, the conditions thought to produce it. Figures from the two kinds of measure are not interchangeable.Full entry in the glossary, satisfaction and intention to stay in surveys by Gallup and in peer-reviewed research, but which way the link runs is unclear. Gallup reports that employees with a best friend at work are more likely to recommend their workplace and less likely to plan to leave, based on panel surveys of US employees.²

Academic work points the same way with broader measures. A 2008 Journal of Applied Psychology meta-analysismeta-analysis: A study that combines the results of earlier studies on the same question into one overall estimate. Pooling makes the estimate more precise, but it cannot repair the studies it pools: a meta-analysis of surveys is still survey evidence.Full entry in the glossary pooled 161 samples and found support for most of the links it proposed between coworker support and employees’ attitudes, withdrawal from work and effectiveness.³ Coworker support is wider than friendship: a helpful colleague you never see after hours counts too.

The catch is direction. Happy, engaged people may make friends more easily, and a well-run team gives friendships room to form, so the friendship may be a sign of a good job as much as its cause. A product manager who loves her work, has a fair boss and a sensible workload will probably report a work friend and high satisfaction, and the survey cannot say which came first.

Most people also manage without close friends at work. Marissa King and Emma Seppälä, writing for UC Berkeley’s Greater Good Science Center, note that some people simply prefer a more formal relationship with colleagues, and that values such as vulnerability, authenticity and compassion can make a workplace feel friendly without deep friendships.⁴ Treat a work friendship as something that often comes with a good job, and judge a workplace on the conditions that let friendships form.

Why Gallup asks about a “best friend at work”

Gallup asks employees whether they have a best friend at work because, it says, its early research found that wording predicted workplace outcomes better than asking about a “friend” or a “good friend.” The item is element 10 of the 12 in Gallup’s engagement survey, the Q12.⁵ Gallup sells that survey to employers as a service.²

We use this very specific wording because early research indicated that having a “best friend” at work was a more powerful predictor of workplace outcomes than simply having a “friend” or even a “good friend.”

The question still draws pushback from managers who see friendship as bad for productivity, a 2018 Gallup article notes.⁶ The figures quoted on news and HR sites mostly come from a few Gallup pages, and they are not all the same kind of evidence.

Myth

Research proves that people with three close friends at work are 96% more satisfied with their lives.

Fact

The 96% line comes from a 2006 interview promoting a book. It compares likelihoods, and the page gives no sample, method or study design.

None of this makes Gallup’s data worthless. Its panel work is weighted to match the US population, and its warnings are sensible: Gallup’s 2018 piece says friendships help most once basics such as clear expectations are in place, and can otherwise turn into gripe sessions.⁶ When you see a best-friend statistic, check three things: which country it covers, whether it was observed or projected, and who sells the survey.

The costs most articles skip: exhaustion and upkeep

Work friendships take energy as well as giving it. The best evidence on the trade-off comes from Jessica Methot and colleagues, who measured both sides in the same people and found that friendships at work were tied to performance through routes that pushed in opposite directions.¹

The study

Moderate evidence

Two studies that counted the gains and the drain of work friends

In a study of 168 insurance company employees, people with more friends among the coworkers they relied on for work received better supervisor ratings of job performance. The same friendships were also linked to more emotional exhaustion, which pulled ratings down, partly offset by a better mood. A second study of 182 restaurant and retail sales employees found the friendships helped through trust and hurt through the difficulty of keeping them going.¹

The authors conclude that these friendships still benefit performance, though the gain is muted somewhat by the personal resources they use up. The two studies covered insurance, restaurant and retail staff, and neither could assign friendships at random, so they show associations rather than proof; both measured friendships with the people respondents worked with, which the authors call multiplex ties.¹

Julianna Pillemer and Nancy Rothbard, in a 2018 Academy of Management Review paper, explain why the strain arises. Friendship is informal, voluntary, based on shared needs and pursued for its own sake, while organizations run on formal roles, relationships you cannot choose, tit-for-tat exchange and business goals.⁸ Rothbard added in an interview that close friendships can become distracting, and that friends can find it hard to challenge each other in a group decision.⁹

A team lead whose closest colleague keeps coming to her desk to talk through a rough patch at home faces exactly this: she wants to help, and her own deadline slips. Emotional exhaustionemotional exhaustion: Feeling drained and used up by work, with little energy left. It is the exhaustion dimension of burnout, one of three, and on its own it is not a diagnosis; persistent exhaustion is worth raising with a doctor.Full entry in the glossary builds when that pattern has no limit, so set a rhythm both of you can keep, such as a walk at lunch rather than an open door all afternoon.

What friends bring

• Trust and empathy
• Support and a better mood
• Easier sharing of know-how

What they can cost

• Time and effort to keep it going
• Emotional exhaustion
• Clashes with roles, rivalry and outsiders

Can work friendships hurt your career?

Work friendships can hurt a career when friendship and business interests collide: when one friend manages the other, when both compete for the same promotion, or when other colleagues feel shut out. Paul Ingram and Xi Zou of Columbia University, in a 2008 review, trace these problems to the mix of affection and business interest in the same relationship.¹⁰

Rivalry is the sharpest case. Ingram and Zou describe two friends on the same team where one is promoted and the other is not; drawing on research on how people protect their self-image, they note that in an area people care about, a friend’s success can feel more threatening than a stranger’s.¹⁰ Talking early about the fact that you may compete, before a decision lands, gives both of you a script for the day one wins.

Cliques are the cost to everyone else. Rothbard and Pillemer describe how a visible circle of close friends can leave colleagues on the outside feeling excluded, and how social media now shows who went out without whom.⁹ If you belong to a tight group, invite others in on purpose and keep work decisions out of the private chat.

Departures carry a cost too. In a 2009 study of a regional bank and a national hospitality firm, coworkers’ job hunting and how rooted they were in the job helped explain who quit, over and above other individual and group-level predictors.¹¹ When a close colleague starts looking elsewhere, notice whether your own restlessness is yours or theirs.

Should you be friends with your manager or direct reports?

Friendship across a reporting line is possible, but the evidence urges caution. Tom Rath argued in Gallup’s 2006 interview that bosses should try to befriend the people they manage.⁷ Pillemer and Rothbard, by contrast, count formal roles among the main sources of friction in workplace friendship.⁸

The two views can be reconciled in practice. Rothbard’s advice for friendships that cross a reporting line is to be scrupulous about fair process: make the criteria for assignments, ratings and pay explicit, because colleagues watching will assume favoritism otherwise.⁹ A manager who plays tennis with one direct report every weekend can protect both of them by writing down the criteria before allocating the next big project, and by handing any rating of that friend to a second reviewer where the company allows it.

How remote and hybrid work changes friendships at work

Remote work can thin out the everyday contact that work friendships grow from. In a study of 61,182 US Microsoft employees in the first half of 2020, Microsoft researchers estimated that the firm-wide shift to remote work left collaboration networks more static and siloed, with fewer links between groups.¹² Most of the authors worked at Microsoft, studying their own firm, and the data cover one technology company.

Gallup reports that a best friend at work has mattered more since the pandemic, and argues that employees without one became more isolated working from home.² Yet in a separate 2024 survey of US employees, reported with its global workplace report, Gallup found that having a best friend at work was among the weakest of its 12 engagement items in predicting lower loneliness; engagement overall mattered more.¹³ Feeling isolated at work overlaps with loneliness as a wider problem, which has causes well beyond the office. On a remote or hybrid team, contact has to be made on purpose: a few minutes of small talk before meetings, a project with someone outside your group, or office days planned around who will be there.

Cultures differ in how much work and friendship mix

How much people mix work and friendship varies with culture. Ingram and Zou argue that business friendships are particularly hard to build in the United States, where a work ethic rooted in Protestant traditions keeps relational and emotional concerns out of work, while in China, business often runs through dense personal networks and the obligations of guanxi.¹⁰

A study of one global retail bank, which used the same structure and policies in every country, reported results that largely supported the idea that coworker ties in its American, Chinese, German and Spanish divisions follow different cultural templates: market, family, law and friendship.¹⁴ King and Seppälä add that going on holiday with a coworker is rare among US workers while close to half of Indian workers have done it, citing survey figures without naming the survey or its method.⁴ A newcomer from Bangalore who joins a team in Frankfurt may read polite distance as coldness, and a German colleague may read invitations to a family wedding as pressure. Ask how colleagues where you work draw the line before you decide the line is wrong.

Getting the gains with fewer costs

The research suggests friendships help most when work basics are sound and when friends agree on how work and friendship fit together. No trial has tested the steps below; they follow from the costs the studies describe.

Nobody can be close to everyone; the research on how many relationships one person can keep up suggests a handful of strong ties is a realistic aim. When a friendship turns sour, the tools change, and the playbook for difficult coworkers sets them out.

Keeping a work friendship healthy

• Agree on a rhythm for personal talk that fits both workloads
• Say out loud that you may compete one day, before it happens
• Keep decisions, ratings and pay talk out of private chats
• Invite colleagues outside your circle into lunches and projects
• If you manage a friend, write the criteria down before you decide

The bottom line

Talk about the lines before anything tests them: how much personal time fits the workday, what happens if you both want the same promotion, and how decisions get made if one of you manages the other. This week, pick one friendship at work and have that conversation, and if you manage the team, write down the criteria behind your next decision before you make it.