Complementary Teams: How to Stop Hiring Clones

Think about the last person you hired who left you walking out of the interview thinking “this is the one.”

Now ask yourself an uncomfortable question: how much of them was you?

How they structured an answer. What gave them energy. How fast they talked. Which examples they picked to shine. If the honest answer is “quite a lot,” you didn’t discover exceptional talent. You looked in a mirror and liked what you saw.

And that feeling never registers as bias. It registers as judgment.

The bias that disguises itself as good instincts

Selection research has spent decades describing what the literature calls the similar-to-me effect: similarity between evaluator and candidate, in values, habits, style, self-presentation, tends to inflate the rating. It isn’t uniform or enormous across every interviewer, and it’s worth saying plainly: there are studies where the relationship is strong and others where it barely shows up. But the direction of the bias is consistent, and it always leans the same way. Nobody hires someone because they were reminded of the person they can least stand.

The most uncomfortable work on this is Lauren Rivera’s (Hiring as Cultural Matching, American Sociological Review, 2012). She interviewed evaluators at elite professional firms and observed real hiring committees. More than half of the evaluators named “cultural fit” the single most important criterion at the interview stage. The problem is what they meant by fit: similarity to themselves in hobbies, background, and personal style. Not compatibility with the company’s values. Resemblance to the person on the other side of the table.

Rivera sums up the finding in a way that stings: employers were looking for candidates they’d enjoy spending time with, and that preference sometimes outweighed expected productivity.

Translated into the room where you work: “great vibe,” “we clicked right away,” “she’s one of us.” Three phrases that sound like evidence and are actually autobiography.

And the bias doesn’t switch off once the offer is signed. It keeps running in promotions, in who gets the visible project, in who gets invited to the meeting where something actually gets decided. A manager who clones himself once builds a team. A manager who clones himself for three years builds a choir.

What a homogeneous team costs you

A team of similar profiles works incredibly well. That is exactly the problem.

They agree fast. Meetings are short. Nobody asks the annoying question. Cohesion is high, the mood is good, and every soft indicator HR looks at comes back green.

Until the team runs into something none of its members is equipped to see.

Because what they share isn’t just strengths. Five people high in conscientiousness and low in openness will execute flawlessly and will never propose the change of direction the market is screaming for. Five people high in openness and low in conscientiousness will have three brilliant ideas a week and zero shipped projects. A team of Executor profiles delivers on time and never questions whether that thing was worth delivering.

Every profile has a blind spot. It’s the price of having a strength. When the whole team shares the profile, the blind spot stops being individual and becomes structural: nobody inside can see it, because seeing it would require a different way of looking.

And homogeneity is self-confirming on top of that. The homogeneous team reads its own speed of agreement as collective intelligence. “We’re aligned.” No: you’re all looking in the same direction, including the wrong one.

Culture fit doesn’t mean culture clone

There’s a confusion here that has already cost a lot of good people their careers, and it’s worth splitting in two.

Culture fit, properly understood, is alignment on values and ways of working. If your company runs on radical transparency and high autonomy, someone who needs permission for everything and hides mistakes will suffer, and will make you suffer. That’s real fit, and it’s measurable: it’s what we work through in the complete culture-fit guide, and it’s why a generic assessment fails to capture your culture.

Culture fit, badly understood, is similarity of personality, style, and biography to whoever’s making the call. That isn’t culture. That’s comfort.

The practical difference is this: values should be shared, behavioral profiles should not. A team can, and should, share the “why” and disagree on the “how.”

That’s where culture add comes from: instead of asking “does this person fit what we already are?”, you ask “what does this person bring that we don’t have today?”. The first question walks you toward a team that looks more and more like itself. The second forces you to know what you’re missing, which is a much harder and much more useful question.

And to answer it you need something most companies don’t have: a map of what they already are.

How to read your team’s composition

You can’t complement what you can’t see. This is the step almost everyone skips: they hire “to add diversity” without ever having measured the current composition, and the result is a new gut call replacing an old gut call.

With comparable profiles, the same instrument applied across the whole team, not a different one per person, the reading is straightforward. It’s the same argument we make in one engine for the whole talent lifecycle: if you measure the candidate with one ruler and the employee with another, the numbers don’t compare and you’re back to deciding by eye.

Three readings you can run with OCEAN+ the same day you have the team’s profiles:

1. Concentration. What percentage of the team shares the dominant profile? If more than half land in the same archetype, you don’t have a team: you have a choir with different names on the org chart.

2. Gaps. Which dimension is systematically low across everyone? A product team where nobody scores high on openness won’t innovate no matter how many OKRs you throw at it. One where nobody scores high on conscientiousness won’t deliver no matter how much talent it has.

3. Distance to the role. The profile an engineering manager needs is not the profile a product manager needs, even though they sit in the same meeting. Composition is read against roles, not in the abstract.

What the evidence actually says (and what it doesn’t)

I want to be careful here, because the “cognitive diversity equals better performance” literature gets cited with far more confidence than it deserves.

Bell’s (2007) meta-analysis on team composition and performance found something that doesn’t get repeated on LinkedIn: what predicts performance most consistently is not the variance of personality inside the team, but the average level of certain traits, conscientiousness, agreeableness, openness, in field settings. The effects of mere personality diversity on performance are small and depend heavily on context and task.

In other words: throwing different profiles together for its own sake does not make you a better team. A diverse, badly managed team performs worse than a homogeneous, well-managed one.

So why keep insisting on complementarity? For two reasons the evidence does support:

• Capability coverage. A team needs someone who sees the detail, someone who sees the system, someone who pushes, and someone who brakes. That’s role composition, not decorative diversity.
• Reducing the collective blind spot. It isn’t a promise of more quarterly productivity. It’s an insurance policy against the big mistake, the one nobody saw coming because nobody was equipped to see it.

Complementarity isn’t a shortcut to performance. It’s a defense against silent failure.

Healthy friction and destructive friction

And here’s the nuance that makes all of this work or blow up.

The meta-analysis by De Wit, Greer and Jehn (2012), across nearly 90 independent samples, found that task conflict, arguing about the what and the how, has, on average, a relationship close to zero with team performance. Neither good nor bad. But it found something more specific and more useful: task conflict stops being harmful when it’s decoupled from relationship conflict. When arguing with the idea gets confused with attacking the person, everything sinks. Relationship conflict is consistently negative, with no nuance.

Translated into team composition:

Healthy friction is an Analyst asking an Entrepreneur for evidence before committing budget. It’s annoying, it’s uncomfortable, it drags the meeting out twenty extra minutes, and it prevents a six-month mistake. It requires both of them to share the goal and the rules of the game.

Destructive friction is the same scene when they don’t share values: then “you asked me for evidence” reads as “you don’t trust me,” and the technical argument turns personal.

That’s why the complementary team has one non-negotiable condition: different profiles, shared values. Without the second, the first is a civil war.

The all-star team of the same type

One last mistake, and it’s the most expensive one because it’s made with the best intentions.

You hire the best. Then you hire another best. And another. And because your definition of “the best” is anchored in one profile, the one who shines most in interviews, the one who argues best, the one who answers fastest, you end up with five stars of the same type competing for the same space.

Five Strategists and no Executor is not a high-performing team. It’s a debate club with a budget.

Team performance is not the sum of individual talent. It’s what’s left after you subtract destructive friction and add coverage. Hiring the same excellent profile five times maximizes the first term and ruins the other two.

What you can do this week

Without buying anything new:

One. Take your most critical team and write down, from memory, each person’s dominant profile. Then count the repeats. If more than half share a profile, you now know why the meetings are so comfortable.

Two. Review the last three people you hired for that team. Ask yourself what each one brought that the team didn’t have. If you can’t answer in one sentence per person, you hired by mirror.

Three. On your next search, write the profile of the gap before you see the first resume. Not “the best possible candidate”: the candidate who covers what nobody covers today. It’s a different sentence, and it produces a different shortlist.

---

A team of clones doesn’t fail on Monday. It fails the day the problem shows up that none of its members was trained to see, and by then you’ve spent three years hiring in the same direction.

The good news is that composition can be measured, and you don’t have to wait for the next hire to start: you start by looking at the people you already have.

Meet Talento Index to see the real composition of your teams, or see how assessing your current team works with the same engine you use to hire.