Execution vs. Interpersonal Capability

This module teaches a finding that is easy to state and easy to misuse.

RESEARCH FINDING In the largest datasets of structured CEO assessments, candidates' rated abilities collapse into a small number of factors, and one of them is a contrast between execution — resoluteness, efficiency, persistence — and interpersonal ability — team orientation, listening, warmth. Candidates strong on interpersonal ability are more likely to be hired; candidates strong on execution are more likely to succeed and to advance (Kaplan, Klebanov & Sørensen, 2012; Kaplan & Sorensen, 2021).

The easy misuse is to conclude that warmth does not matter and that the ideal CEO is a demanding executor who ignores people. That is not what the evidence says, and the second half of the library — on humility, top-team integration, psychological safety and employee silence — explains why: execution runs on information and on a team that stays, and both are interpersonal products (Ou et al., 2014; 2018; Edmondson, 1999; Milliken, Morrison & Hewlin, 2003).

The correct lesson is about interaction. Execution and interpersonal capability multiply rather than add. A zero in either collapses the product — the likable CEO who never decides, the demanding CEO whose team goes quiet and then leaves.

In the Effectiveness Equation, this module works on the Behaviors term: what a CEO actually does with the traits they have, and which behaviors the evidence links to results rather than to being chosen.

Boards hire the person they like in the room; results come from the person who finishes — and finishing, at scale, requires a team that will still tell you the truth.

The execution-versus-interpersonal contrast is real, measurable and predictive. But it describes a tension within a single job, not a choice between two kinds of CEO. The mature CEO is demanding about outcomes and disciplined about the relationships that supply the information outcomes depend on.

The structured-assessment evidence

RESEARCH FINDING Kaplan, Klebanov and Sørensen (2012) analyzed detailed assessments by one professional firm of 316 CEO candidates at 224 companies involved in private-equity and venture-capital transactions between 2000 and 2006, rated on more than thirty abilities. Factor analysis produced two main dimensions: general ability, and a contrast between execution or "hard" abilities and interpersonal or "soft," team-oriented abilities. Both buyout and VC investors tended to hire CEOs high on general and execution ability. Subsequent success — measured by subjective investor ratings and objective outcomes — was more strongly related to execution abilities than to interpersonal ones. Incumbency was only marginally related to performance once measured ability was controlled.

What this can support: among PE/VC-backed CEO candidates, execution abilities predicted success more strongly than interpersonal abilities. What it cannot support: a claim about public-company CEOs (the sample is not representative), a causal claim (associations only), or a claim that interpersonal ability is irrelevant — it was less predictive, not unpredictive. Outcomes are partly subjective investor ratings, and all ratings come from a single assessor.

RESEARCH FINDING Kaplan and Sorensen (2021) extended the dataset to 2,603 assessments of candidates for CEO, CFO, COO and other roles between 2000 and 2013. Four factors explained over half the variance: general ability; execution versus interpersonal; charisma versus analytical; strategic versus managerial. CEO candidates scored higher on all four. Candidates with stronger interpersonal skills were more likely to be hired, yet the factors — including execution — predicted later advancement to CEO. The authors suggest boards may over-weight interpersonal skills at hiring. The pattern held across public, PE-backed and VC-backed firms.

What this can support: a hiring bias toward interpersonal skill that is not matched by later advancement. What it cannot support: any firm-performance claim; the outcome is career progression.

The parallel in general leadership research

RESEARCH FINDING Judge, Bono, Ilies and Gerhardt (2002) found across 73 general (non-CEO) samples that agreeableness correlated .05 with leadership emergence but .21 with leadership effectiveness, while conscientiousness — the trait of persistence and discipline — correlated .33 with emergence but .16 with effectiveness. The constructs differ from Kaplan's factors, so this is not a replication, but it is a related pattern: what gets a person chosen and what makes them effective are different lists.

Why interpersonal capability still matters: the counterweight

RESEARCH FINDING Ou, Tsui, Kinicki, Waldman, Xiao and Song (2014), using data from 328 top-team members and 645 middle managers in 63 private Chinese companies at two time points, found CEO humility — self-awareness, openness to feedback, appreciation of others, low self-focus — positively associated with empowering leadership behaviors, which related to greater top-management-team integration, which related to an empowering climate perceived by middle managers, which was associated with their engagement, commitment and job performance. Ou, Waldman and Peterson (2018), in 105 small-to-medium US software and hardware firms, found CEOs rated more humble by their top teams had more integrated teams and smaller CEO–team pay gaps, which were in turn linked to a more ambidextrous strategic orientation and better firm performance.

What these can support: CEO humility, as rated by others, is associated with better-integrated top teams and, through them, with organizational outcomes. What they cannot support: generalization to large public firms (both samples are SMEs), or a causal claim (correlational with time separation). The humility-to-financial-performance link is weaker than the humility-to-team-integration link; Module 5 treats this honestly.

RESEARCH FINDING Owens and Hekman (2012), from 55 interviews, identified three humble-leader behaviors — acknowledging limits and mistakes, spotlighting followers' strengths, and modeling teachability — and noted humility is less effective under extreme threat or time pressure. Owens and Hekman (2016), across three studies with 607 participants in 161 teams including an experiment, found leader humility spread through teams and improved team performance. These are team-level, not CEO-level, findings.

RESEARCH FINDING Edmondson (1999), in 51 teams at one manufacturer, found team psychological safety associated with learning behavior and, through it, performance, with team-leader coaching as an antecedent. Milliken, Morrison and Hewlin (2003) found 85% of 40 interviewed employees could recall an occasion when they felt unable to raise an important issue with a superior, most often from fear of being labeled negatively or of futility. Detert and Edmondson (2011), across four studies, identified taken-for-granted "implicit voice theories" that suppress upward candor even where the environment is objectively safe. None of these measure CEO information environments directly; the inference to CEOs is an extrapolation.

Behavior at scale

RESEARCH FINDING Bandiera, Prat, Hansen and Sadun (2020), using time-use diaries for 1,114 manufacturing CEOs in six countries, classified CEO behavior on a spectrum from "manager" (more one-on-one meetings with operational staff, plant visits) to "leader" (more multi-function, multi-participant meetings with C-suite executives). A one-standard-deviation move toward leader behavior was associated with about 7% higher sales, with the difference emerging only about three years after appointment. The authors estimate 17% of firms had a mismatched CEO type, and explicitly caution that this is a matching story — not evidence that leaders are always better than managers.

Where the evidence is weak

The execution finding rests on one assessor's ratings of PE/VC candidates with partly subjective outcomes. The humility findings come from Chinese private firms and US tech SMEs. The silence findings come from non-CEO samples. Nobody has run a study in which execution and interpersonal capability are measured on the same public-company CEOs and their interaction tested against long-run performance. Gow et al. (2016; working paper) found extraversion negatively related to ROA, which is suggestive but measures a trait, not interpersonal skill. Herrmann and Nadkarni (2014) found agreeableness related to both initiating and implementing change in 120 SMEs — a hint of interaction, but from small firms. The interaction claim in this module is INTERPRETATION built from studies that each support one side.

Three factors, and what they measure

FACT The structured assessments behind Kaplan's studies rate candidates on thirty-odd specific abilities through multi-hour interviews. Statistically, those ratings collapse into a few dimensions. Three matter here.

General ability is the first factor in both studies: candidates who rate well tend to rate well on most things. It is closest to "how strong is this person overall," and it is not the interesting part.

Execution — the "hard" pole of the second factor — includes resoluteness, efficiency and persistence. It is the capacity to decide, to drive, to hold people to commitments, to finish. Kaplan, Klebanov and Sørensen (2012) describe it as the pole more strongly related to success.

Interpersonal — the "soft" pole — includes team orientation, listening, respect for others and openness to criticism. It is the capacity to make people want to work with you and to keep them working together.

INTERPRETATION The factor is a contrast, which means the assessment data found candidates tend to be stronger on one pole than the other. That is a fact about the population, not a law of nature. The rare candidate high on both exists; the point of this module is to explain why that candidate is worth finding.

Why boards over-weight the interpersonal pole

Hiring is a social process. A board meets a candidate three or four times in rooms designed to be pleasant. Interpersonal ability is fully visible in that setting; execution ability is mostly invisible — it shows in what the candidate did over years, which the board reads in a résumé and a reference call. The board also emerges from each meeting with a feeling, and the feeling is produced by the interpersonal pole.

This is Module 1's emergence-versus-effectiveness gap operating at the top of the pyramid. RESEARCH FINDING Judge et al. (2002) found the traits that get people chosen as leaders differ from those that make them effective. RESEARCH FINDING Kaplan and Sorensen (2021) found the same thing for CEO hiring: interpersonally strong candidates get chosen; execution-strong candidates advance.

INTERPRETATION There is a second reason boards over-weight interpersonal ability, and it is rational: boards have to work with the CEO. A board choosing someone pleasant to deal with is optimizing a real cost — its own. The problem is that the board's experience of the CEO is not the company's.

"Likable but ineffective"

The pattern: a CEO everyone enjoys, who listens well, builds consensus, avoids conflict, and whose company drifts. Decisions get made slowly or not at all. Under-performers stay because removing them would be unkind. Strategy is a synthesis of everyone's preferences. The board likes the CEO and cannot quite say what is wrong.

INTERPRETATION The mechanism is not the presence of warmth but the absence of execution. On the Effectiveness Equation, the Behaviors term has a near-zero in it: nothing is finishing. The CEO's interpersonal skill is real and is actively masking the problem, because a company run by a likable CEO feels healthy long after it has stopped performing.

Diagnosis: ask what the CEO has stopped — a product, a person, a project — in the last year. Ask how long the three weakest executives have been in role. Ask whether the CEO has ever said the first sentence of the Two-Sentence Test ("We're going to do this") in a form that cost anyone anything.

"Demanding but effective"

The pattern: a CEO who decides fast, holds people to numbers, tolerates no excuses, and gets results. Turnover is high, sometimes very high. The people who leave are often the most independent. Those who stay learn to bring good news. The board sees the numbers and tolerates the churn.

INTERPRETATION This CEO has the execution pole and is missing the interpersonal pole, and for a while it works, because execution is what the results depend on. But the missing pole is not a nicety. It is the information system. The silence research is explicit: most employees can recall withholding an important issue from a superior, mainly from fear of being labeled negatively (Milliken et al., 2003), and the rules that suppress voice persist even in objectively safe settings (Detert & Edmondson, 2011). A demanding CEO does not have an objectively safe setting. The team is not lying; it is filtering. And Ou et al. (2014; 2018) show the top-team integration that a humble CEO produces is the channel through which the organization's information and effort actually flow.

Diagnosis: ask when the CEO last changed a decision because of something a subordinate said. Ask who left in the last two years and what they said on the way out. Ask whether the numbers are being reported or produced.

How the two interact

FRAMEWORK Execution and interpersonal capability are multiplicative in the Behaviors term, for three reasons.

Execution needs information. A CEO cannot finish the right thing without knowing what is happening, and at scale what is happening arrives only through people who choose to tell you. Interpersonal capability is what makes them choose to. This is why Module 5 treats humility as an information-quality tool.

Execution needs a team that stays. Ou et al. (2018) link CEO humility to top-team integration and smaller pay gaps, and those to ambidexterity — the ability to exploit and explore at once. A CEO who executes alone can do one thing at a time. A CEO with an integrated team can do several.

Interpersonal capability needs something to coordinate. Bandiera et al. (2020) found "leader" behavior — multi-function meetings with the C-suite — associated with higher sales after about three years, while cautioning that this is a matching story. INTERPRETATION Coordination pays only where there is execution to coordinate. A likable CEO holding excellent cross-functional meetings about a strategy nobody will drive is producing consensus without motion.

The Player → Coach → Architect model (Module 8) adds a scale dimension. HYPOTHESIS At Player scale, a founder's personal execution can carry the company and interpersonal gaps are survivable. At Architect scale, the CEO executes almost nothing directly; results come through the team and the information system, and the interpersonal pole becomes the mechanism of execution rather than an alternative to it. The demanding-but-effective CEO tends to fail at the scale where they can no longer see the work themselves.

What "interpersonal" does not mean

INTERPRETATION Interpersonal capability in the Kaplan sense is not niceness, agreeableness or conflict avoidance. It is the set of behaviors that make people willing to work with you and tell you the truth: listening, respect, openness to criticism, giving credit. Owens and Hekman's (2012) three humble-leader behaviors — admitting limits, spotlighting others' strengths, modeling teachability — are a good operational definition. None of them requires being soft about outcomes. A CEO can be entirely demanding about results and entirely humble about their own view of how to get there. That combination is what the second sentence of the Two-Sentence Test sounds like in practice.

This is a fictional composite.

Larkspur Advisory is an engineering consultancy: $75 million in fees, 380 staff, eleven equity partners, founded in 1994. Owen Marsh became managing partner in 2017 by a unanimous vote. He was the partner everyone wanted in the room with a client: warm, unhurried, genuinely interested in people, incapable of a harsh word. Under him, Larkspur's culture surveys were the best in its peer group.

By 2023, revenue per professional had fallen 14% in real terms. Two practice areas had been unprofitable for three years; both were led by partners Marsh had known for twenty years. A planned move into infrastructure advisory had been "in development" since 2019. Three of the firm's strongest under-forties had left for competitors, each telling Marsh in their exit conversation that they loved the firm and could not see it changing. He agreed with them.

The partners' council, prompted by the firm's bank, commissioned an outside review. The reviewer's summary was blunt: Larkspur had a managing partner with exceptional interpersonal capability and no record of finishing a hard decision. She asked Marsh what he had stopped in six years. He could not name anything.

Marsh stepped aside, graciously, in early 2024. The council appointed Jo Tanaka, a partner from the transport practice who had built it from $6 million to $19 million with a reputation for exacting standards and short meetings. In her first year she closed one unprofitable practice, restructured the other, launched infrastructure advisory with a hired team, and moved revenue per professional up 9%. She also lost two partners and a fifth of the associates in the closed practice, and the firm's culture score dropped to the bottom of its peer group.

In her second year the trouble started. A large client's project ran over budget for four months before Tanaka heard about it; the project lead later said he had "assumed she knew and would come down on him anyway." The infrastructure team's head resigned, citing "no room to disagree." The council chair — who had backed her — asked her to meet Marsh.

What Marsh told her was not to be nicer. It was to notice that the people around her had stopped bringing her problems, and that her execution was now running on stale information. She started two practices: a standing monthly session where each practice head had to bring the thing they least wanted to tell her, and a personal rule of stating what she thought and then asking what she was missing before deciding. Neither made her less demanding. Within a year, the project-overrun problem had not recurred, and the firm's culture score recovered to mid-pack while revenue per professional continued to climb.

INTERPRETATION Larkspur had a likable-but-ineffective managing partner followed by a demanding-but-effective one, and it needed the second. But the second was effective only until her missing pole cut off the information her execution ran on. The fix was not warmth; it was the interpersonal behaviors that keep the truth flowing.

Put four archetypes into Larkspur's managing-partner chair in 2024.

The Turnaround CEO does what Tanaka did, faster and harder. Closes both practices, exits four partners, resets compensation. Gain: the economics fix in eighteen months instead of thirty. Cost: a partnership is not a corporation; the survivors own the firm and remember. The Turnaround CEO's execution pole is exactly right for year one and the interpersonal gap becomes the problem by year two, as it did for Tanaka — but sooner, because a partnership's information flows are relational.

The Professional Services CEO — a partner who has run a large firm before — starts with the partners. Builds a case, negotiates the practice closures as retirements, and spends a year on consensus before acting. Gain: the firm stays whole. Cost: another year of drift, and the risk that consensus produces Marsh's outcome under a different name. Whether this works depends on whether the interpersonal work is a route to execution or a substitute for it.

The Scale-Up CEO treats Larkspur as a growth problem: hires the infrastructure team, invests ahead of revenue, lets the unprofitable practices shrink by neglect. Gain: motion, and a story the under-forties want to stay for. Cost: a levered partnership cannot fund growth and carry losses at once, and neglect is not a decision. This CEO has execution on the initiation side and none on the stopping side.

The Operator CEO runs the numbers, sets utilization targets by practice, and manages to them. Gain: the drift stops, measurably. Cost: an engineering consultancy's value is in judgment and relationships, not utilization, and the Operator may execute the firm into a body shop. The interpersonal pole here is not about the CEO's warmth but about understanding what the firm's people actually produce.

INTERPRETATION Larkspur's problem was a missing execution pole. Any of the four would have supplied it. The differentiator is which one would also preserve the interpersonal machinery — partner trust, upward candor — that a consultancy runs on. The Professional Services CEO understands this best and executes slowest; the Turnaround CEO executes fastest and understands it least.

Each pole has its own rungs on the Overuse Ladder.

FRAMEWORK Execution overused. Decisiveness → impulsiveness: decisions made before the information arrives, because waiting feels like weakness. Persistence → stubbornness: the commitment that finished the last project keeps the wrong one alive. Dominance → intimidation: the intensity that drove results now suppresses the dissent that would have caught the error. The end state is a CEO running a company on filtered information, at speed, with confidence. Chatterjee and Hambrick (2011) found narcissistic CEOs discount objective feedback; an intimidating executor achieves the same effect by making sure the feedback never arrives.

Interpersonal overused. Empathy → conflict avoidance: the CEO's care for individuals prevents decisions that would hurt any of them. Humility → hesitation: openness to others' views becomes an inability to hold one's own. Consensus becomes the decision process and drift becomes the strategy. Owens and Hekman (2012) note humility is less effective under extreme threat or time pressure; a company in trouble led by a CEO who over-uses the interpersonal pole gets slower exactly when it needs speed.

Early warning signs a CEO or board could notice:

• Execution overuse: senior turnover concentrated among the independent-minded; the CEO is rarely surprised by bad news because bad news rarely arrives; exit interviews mention "no room to disagree"; the CEO has not changed a decision on a subordinate's input in a year.
• Interpersonal overuse: nothing has been stopped; the weakest executives have long tenures; strategy documents are lists of everyone's priorities; the CEO's own view on the hardest question is not written down anywhere.
• Both: the board's opinion of the CEO is formed mostly in board meetings, and the board has no independent channel to the top team.

The correction runs in opposite directions but uses the same instrument: the Two-Sentence Test. The over-executor must practice the second sentence. The over-relator must practice the first.

1. On the execution–interpersonal contrast, which pole are you stronger on? Now: which pole did the people who hired or promoted you see?
2. Name the last decision you changed because a subordinate told you something you did not want to hear. If it is more than six months ago, what does that suggest?
3. Name the last thing you stopped — a product, a project, a person — that cost you a relationship. If you cannot, which pole is missing?
4. Who has left your organization in the last two years because of how you lead, rather than what you decided? What would they say you did not hear?
5. Write down the three things your top team is least likely to tell you. How do you know they are not already true?
6. If your board could observe you only through your top team's eyes for a quarter, how would their assessment of you change — and in which direction?

• Structured assessments of thousands of executives reduce to a few factors, one of which contrasts execution (resoluteness, efficiency, persistence) with interpersonal ability; execution predicts success and advancement more strongly (Kaplan, Klebanov & Sørensen, 2012; Kaplan & Sorensen, 2021).
• Boards appear to over-weight interpersonal skill at hiring because it is visible in the room and because boards must work with the CEO; execution is visible only in the record.
• "Likable but ineffective" is a missing execution pole masked by warmth; "demanding but effective" is a missing interpersonal pole masked by results — and it fails when the team stops supplying information (Milliken et al., 2003; Detert & Edmondson, 2011).
• Execution and interpersonal capability multiply: execution runs on information and a team that stays, both of which humility and integration produce (Ou et al., 2014; 2018); interpersonal capability without execution produces consensus without motion.
• Interpersonal capability is not niceness. Admitting limits, crediting others and modeling teachability (Owens & Hekman, 2012) are fully compatible with being demanding about outcomes — that combination is what the Two-Sentence CEO sounds like.