High performer spillover: who you sit next to matters

Sam came to see me after I’d moved her desk.

I’d sat her next to one of the developers, and she had a confession. She hadn’t liked him. She was quite apologetic about it, and felt a bit silly saying so now, because he’d made her a cup of tea every day and she’d discovered she rather liked him after all.

Nothing especially dramatic had happened. She’d just spent enough time beside him to change her mind.

But she would never have chosen that seat. Left to herself, she’d have stayed on the other side of the office, perfectly comfortable with an opinion she had no reason to question.

That was part of why I moved desks every quarter at Peer 1, a habit I also used at Rackspace and IT Lab. Left alone, the techies kept to their corner and sales sat with sales. After a few moves, the business began to feel more joined up.

Think about what you usually do when two teams won’t talk to each other. Book some training? Organise an away day? Spend a small fortune getting everyone to collaborate somewhere other than the place where they actually work?

Moving a desk costs nothing.

And sometimes the thing standing between two people is simply that neither has ever had to get to know the other.

What is high performer spillover, and how far does it reach?

There’s a performance argument for this too.

High performer spillover is the effect someone has on the work of the people around them. Put a strong performer nearby and others can improve. Put someone with harmful habits in the same position and those habits can spread instead.

The Kellogg workplace design research found that people working within about 25 feet of a high performer got better at their own jobs. Arrange the seating well and organisational performance rose by roughly 15%, worth about $1 million a year in a business of around 2,000 people.

That’s a sizeable return for changing where people sit.

But there’s a catch. The effect weakened with distance, disappearing beyond roughly 25 feet. It also faded within about one to two months after the strong neighbour left.

So this isn’t quite the same as sending someone on a course and hoping they come back with a skill they’ll keep forever. Who they work beside is part of what helps them perform.

I’ve watched the same thing in the businesses I coach. A high-performing individual brings something to the people around them, not just to their own work. Whether you make use of that depends partly on who gets to be around them.

Who should sit next to whom?

You might assume the answer is to put your best people together.

The research suggests something more interesting. The biggest gains came from complementary pairings: someone fast beside someone careful, for example. Productivity rose by about 13% and effectiveness by about 17%. Each person retained their strength while improving on their weaker side.

In other words, you didn’t have to sacrifice speed to get better quality.

Now look at your office. Is that how you’ve arranged it?

Or have you put all the engineers together because that seems efficient, and all the closers on one row because they understand each other? It feels natural enough. But it can leave speed as one team’s concern and quality as another’s, with work travelling back and forth between them.

There is a qualification here. For open-ended skills such as creativity, the research suggests that similar people sitting together can still be useful. Complementary neighbours were particularly valuable for finite skills such as speed, accuracy and throughput.

So the point isn’t to shuffle everyone at random. It’s to think about what you want each person to pick up from the person beside them.

You can invest heavily in talent and still miss this. If all your A-Players sit in one corner, you’re limiting who benefits from having them there.

Why people sort themselves into sameness

Of course, if complementary neighbours are so useful, you might wonder why people don’t choose them.

Because choosing a neighbour rarely starts with, “Who would help me improve?”

It starts with, “Who do I like?”

A study of university friendships makes the problem rather clear. Angela Bahns, Kate Pickett and Chris Crandall approached pairs of friends at a large Midwestern state university with more than 25,000 students, and at small colleges averaging about 500.

On nearly every attitude, belief and health behaviour they measured, friends at the big university were more alike.

You’d think a bigger, more diverse campus would expose you to a wider range of people. And it does. But it also gives you enough choice to find someone remarkably similar to yourself.

At the small colleges, that choice was more limited. Friends were less alike, yet they rated their friendships as closer. They’d been friends for about as long as the big-campus pairs, so this wasn’t simply a matter of having known each other longer.

The authors describe the irony in Group Processes & Intergroup Relations: more diversity on campus, less diversity within each friendship.

Your office can work in much the same way. Give everyone complete freedom to choose where they sit and you may get comfortable little groups who understand one another very well, while continuing to misunderstand everyone else.

Moving desks interrupts that. It gives people a reason to discover something their first impression missed.

Sam didn’t need to become more open-minded in the abstract. She needed to sit next to the developer.

Why toxic spillover costs more than A-Player spillover pays back

There is, however, a fairly important condition attached to all this.

You need to be careful who you’re spreading around.

If good performance can influence a neighbour, so can harmful behaviour. And the research suggests the damage from the latter can outweigh the benefit from the former.

The companion Toxic Workers research found that replacing one toxic worker with an average performer was worth more than twice the value of hiring a superstar.

Source: Kellogg Insight’s companion article on toxic workers.

That calculation came before legal costs, reputational damage or the loss of good people who decide they’ve had enough.

In the workplace design research, roughly 2% of workers were fired for harming people or property. A one-standard-deviation increase in the concentration of toxic workers nearby raised a neighbour’s likelihood of later being fired for toxicity by about 27%.

And while the positive effect was concentrated within about 25 feet, the toxic effect could extend across a whole floor. It appeared in the month of exposure and largely disappeared within about a month of the person’s removal.

You can see the same tendency in a much smaller setting.

Will Felps at the University of Washington put students into teams of four, with one member secretly played by an actor behaving as a jerk, a slacker or a pessimist. Each session lasted about 45 minutes. Across dozens of trials, the teams with the actor performed 30 to 40% worse than those without him.

The others also began copying his behaviour.

Felps described the experiment in a This American Life interview. His review with Terence Mitchell and Eliza Byington found a similar pattern across workplace teams: one negative member can spoil the group, and one or two good workers can’t simply cancel out the damage.

Now, a distinction matters here. In the toxic-workers paper, “toxic” means serious misconduct leading to dismissal. That’s narrower than what I call a Toxic A-Player: someone who hits the number and poisons the room.

You shouldn’t confuse the two definitions. But neither gives you a reason to leave harmful behaviour at the centre of the business.

When my clients first measure their A-Player share, it usually falls somewhere between 10 and 35%, and there’s nearly always a Toxic A or two. With relatively few A-Players and a couple of loud Toxic A’s, you can see how the wrong habits become the ones everyone notices.

This is where a well-intentioned desk move can make things worse.

If you already know someone is poisoning the room, giving them new neighbours isn’t a culture initiative. You’re just introducing more people to the problem.

Doesn’t moving people around just annoy everyone?

Yes. Everybody always bitched and moaned about it.

And some of the objections were perfectly reasonable. Moving is disruptive. Sitting beside someone new can interfere with deep work. Your best engineers may have good reasons to want to sit together, particularly on creative work, where the research offers some support for keeping similar people close.

I’ll grant all of that.

I’d still move them.

You have to weigh the disruption you can see against the cost of leaving things as they are. A desk move produces immediate complaints. Two departments quietly misunderstanding each other for three years may produce no complaints at all. You just get used to working around it.

In my experience, the moaning settled after a few cycles. The connections remained worth making.

If deep focus is the concern, protect time for it. If a particular piece of work needs certain people together, account for that. You can take those needs seriously without treating the entire floor plan as untouchable.

Otherwise, Sam keeps her seat. And her opinion.

If the lift fades, why bother moving desks?

At first glance, the fading effect looks like an argument against bothering.

If someone improves beside an A-Player, then drifts back when the A-Player leaves, what have you really achieved?

Quite a lot, provided you understand what’s doing the work.

The workplace design study points to peer pressure and inspiration rather than lasting learning. The positive effects appeared quickly and faded within about one to two months after exposure ended. The neighbour’s presence mattered.

That means you can’t arrange one clever seating plan, tick the box and assume you’ve permanently improved everyone.

It also means people need enough time together for something to happen. Sam came round after about a month. Move her again after a week and she might never have got past her original impression.

So there are two things to manage: enough continuity for relationships to develop, and enough movement for those relationships to extend beyond the same small group.

That was the value of the quarterly habit.

I talk about building the neural networks of a business. Peer coaching creates connections, and seating helps decide who gets connected to whom. Over time, more of the business becomes familiar to more of the people in it.

There’s a related example at a larger scale. Moira Clark at Henley worked on a project with a high-street bank in Sheffield. Moving the manager from the best branch into the worst changed nothing. Transplanting about 30% of the best team tipped the branch.

One impressive person and a different working environment are not necessarily the same thing.

And once you start thinking this way, it’s worth looking beyond desks. Who do you turn to when there’s a difficult decision to make, and whose opinion do you rarely hear? Who regularly works with your strongest people? Who has been left outside those conversations?

Your leadership team has a seating plan too.

What about remote-only teams?

If you’re remote, you can’t move someone’s desk. But you can change who they work with.

Start by looking at the last few weeks. Has everyone worked with the same small group, asked the same people for help and heard the same opinions? If so, you have the same problem as an office where nobody ever changes seats.

Give people a reason to work with someone different. Pair a quick producer with a careful reviewer on a real piece of work. Ask someone from sales and someone from delivery to solve a customer problem together. Let a less experienced colleague work alongside one of your strongest people, rather than only seeing the finished result.

The work matters. Adding someone to another channel or booking a virtual coffee doesn’t give them much opportunity to see how a colleague thinks, handles a problem or gets something done.

Each quarter, look at who has worked together and who hasn’t. Change a few pairings, then give them enough time to produce something together. You’re trying to build working relationships, not fill calendars.

And if someone consistently undermines colleagues on calls or in messages, deal with it. Working from home doesn’t stop other people picking up their habits.

What should a founder-CEO do with this on Monday?

Start with the people who actually spend time together, rather than the tidy boxes on your organisation chart.

1. Draw the real seating plan. Include desks, office-day groups, Slack channels, pairing calls and the room where decisions get made. You’re trying to see who regularly encounters whom.
2. Look for complementary strengths. Name two or three measurable strengths in each team. Where the work depends on speed, accuracy or throughput, consider who could benefit from working beside someone with a different strength.
3. Make movement a habit. Use a quarterly review to create new connections, physically or digitally. Allow enough time for people to get beyond introductions before moving them again.
4. Deal with harmful behaviour first. Don’t use a desk move to postpone a fair dismissal. Training budgets won’t fix a Toxic A, and moving one around gives more people exposure to the behaviour you need to address.

None of this needs to begin with an office redesign. It begins with looking at the arrangement you already have and asking what it encourages.

You already know who’s sitting next to your Toxic A

You could probably draw much of that plan from memory.

You know who sits beside your Toxic A. You may already have noticed someone picking up their habits. You also know which people you’d hesitate to put together.

That last group is worth another look.

Perhaps you have a good reason. Or perhaps you’re protecting a first impression that neither person has ever had to test.

Sam had her opinion of the developer. A month and a daily cup of tea gave her a better one.

Who in your business has been written off, by a colleague or by you, simply because nobody ever moved the desk?

Frequently asked questions

Does sitting near a high performer improve your work?

It can. In the Kellogg workplace design study of roughly 2,000 tech workers, well-arranged, complementary seating lifted organisational performance by about 15%, worth roughly $1 million a year at that scale. Pairing people whose strengths complemented each other raised productivity by about 13% and effectiveness by about 17%.

The effect was concentrated within about 25 feet. It appeared to come more from peer pressure and inspiration than lasting skill transfer, fading within about one to two months after the strong neighbour left.

How close do you need to be for high performer spillover?

About 25 feet was the rough radius in the workplace design research, with the effect weakening as distance increased. In the Kellogg workplace design study of around 2,000 tech workers, that was the range where sitting near a high performer helped most, and the benefit faded within about one to two months once the strong neighbour moved away.

The negative effect could reach further. Toxic behaviour could influence people across a whole floor, not just at the next desk.

For remote teams, there isn’t an equivalent distance established by that study. The useful question is who regularly shares channels, pairing sessions and decision calls.

Do toxic employees really hurt profits more than superstars help?

In the companion toxic-workers research, replacing a toxic worker with an average performer saved about $12,800, compared with roughly $5,300 added by a top 1% superstar. That was before legal costs and reputational damage. In other words, getting rid of one toxic worker was worth more than twice as much as hiring a superstar.

The workplace design research also found that a one-standard-deviation increase in nearby toxic-worker density raised a neighbour’s likelihood of later being fired for toxicity by about 27%.

The research definition concerned serious misconduct. My term Toxic A-Player is broader: someone who delivers results while damaging the people around them. Strong numbers don’t make that damage acceptable.

Does this still matter for remote-only or hybrid teams?

Yes, although the physical-distance findings shouldn’t be transferred directly to remote work. The Kellogg workplace design study measured people sitting within about 25 feet of each other in an office, not colleagues working apart. What still carries over is the principle that the people you work closely with shape your own performance, for better or worse.

Engineers in the same building received about 22% more online feedback than distant teammates while offices were open. That advantage largely disappeared when offices closed.

Microsoft and Berkeley Haas research on roughly 61,000 employees also found that fully remote work made collaboration networks more siloed. Who works with whom still deserves deliberate attention.

How is this different from talent density?

Talent density concerns the share of A-Players in the roles that matter, particularly critical roles, without tolerating C-Players or Toxic A’s. It tells you who you have. High performer spillover is about where you put them, because sitting near a high performer can improve your own work.

High performer spillover concerns the effect those people have on the colleagues around them. You can have strong talent density but limit its wider benefit by keeping all your best people together.

And because the benefit fades when the strong neighbour leaves, it needs maintaining. You can’t arrange it once and assume the job is done.

What should you do next?

---

1. Book a call. If you’re tolerating a Toxic A or struggling to build enough A-Player density, Dominic can help you work out whether the constraint is your people, the design of your leadership team or the way decisions get made. No obligation, no pitch.
2. Grab the book. Mind Your F**king Business covers talent density, Toxic A’s and building a company that doesn’t need you in every room.
3. Read how to deal with Toxic A-Players. Start here if you already know who is poisoning the room.
4. Subscribe to the newsletter. More on talent density, working relationships and building a leadership team that can run things without you.