The Confidence Trap: How Companies Misjudge Talent—and Lose Their Best Leaders

The corporate world has a measurement problem. It’s not showing up in quarterly earnings, at least not yet. But it’s quietly distorting one of the most consequential decisions companies make: who gets promoted.

The pattern is stark. For every 100 men promoted to first-time manager, only 81 women move up, according to McKinsey’s latest Women in the Workplace report. This ‘broken rung’ is where the gap begins and later widens as careers progress. Women of color face even steeper odds. The explanation isn’t what most executives assume. It’s not a pipeline problem or a confidence deficit. It’s a measurement failure, which is costing companies their best talent. And the economic cost is real: research shows that less diverse leadership teams make weaker strategic decisions and correlate with lower innovation and reduced long-term financial performance. In other words, the broken rung doesn’t just limit careers; it drags on competitiveness.

When economist Alan Benson and his colleagues analyzed promotion data from nearly 30,000 management-track employees in 2024, they uncovered something striking: women consistently outperformed men in their current roles, yet received lower ratings for “potential.” More damning still, those potential ratings didn’t predict future success. Women labeled as having less potential went on to outperform male colleagues with identical scores. Benson’s team defined future success using measurable outcomes—subsequent performance ratings, bonus results and promotion velocity. On all three indicators, women labeled as “lower potential” went on to outperform male colleagues with identical ratings.

The problem compounds in how feedback gets delivered. Large-scale text analysis conducted in 2024 by Textio, a company that studies workplace language patterns, shows that women’s performance reviews disproportionately focus on personality traits and labels, such as “abrasive,” “too nice” and “emotional,” rather than on business impact. Men’s reviews, by contrast, center on business outcomes and technical skills. This creates what researcher Abigail Player and her colleagues documented in a 2019 study: men advance on what they could do; women must prove what they have already done. “It’s not a fairness issue alone,” says one tech executive I spoke with. “It’s a capital-allocation failure. We’re underfunding the operators who actually deliver results.”

The root cause? Companies confuse confidence with competence. In promotion discussions, factors like ”executive presence” and “gravitas” carry outsized weight, masking bias and rewarding self-promotion over substance. Ironically, calibration meetings, designed to standardize ratings, can amplify this dynamic. Without structured criteria, confident storytelling trumps comparable results. The loudest voice in the room isn’t necessarily the most capable leader. High performers vote with their feet when they realize the system rewards storytelling over results. The legal and reputational risks of gendered review language only add to the fallout

There’s a cruel irony here. The higher scrutiny women face can push some to invest more heavily in development, which may explain why senior women often rate themselves as highly effective later in their careers. But treating rigorous vetting as a feature, not a bug, is poor governance and even poorer business strategy.

How can we fix the problem?

A growing number of companies are rejecting vibes-based promotion in favor of evidence-based advancement. Their approach: treat potential as a hypothesis that requires proof, not a halo that justifies advancement.

• Define potential concretely. Instead of vague “leadership qualities,” forward-thinking firms specify measurable competencies: learning agility demonstrated through cross-functional projects, decision quality under uncertainty backed by post-mortems and enterprise thinking evidenced by peer and customer feedback.
• Audit the ratings that gate opportunity. Run quarterly audits comparing potential scores to performance ratings, broken down by gender and race. If women systematically score lower on potential despite equal or superior performance, your system, not your talent, needs fixing.
• Replace confidence tests with readiness trials. Rather than relying on gut feelings about who’s “ready,” pilot 90-day live assignments: a customer rescue, a product launch, a cost-reduction program. Establish success metrics upfront and let outcomes speak.
• Ban trait-only feedback in calibration. Prohibit adjectives like “not confident enough” unless tied to specific business impact. Structure promotion discussions like investment committees: require evidence, not impressions.
• Reframe the opportunity itself. Women don’t reject promotions due to insufficient confidence; they decline when the role seems poorly designed or unsupported. Spell out scope, resources, success metrics and structured onboarding. Assign a sponsor and provide a first-90-day playbook.
• Monitor the language. Use tools that flag gendered patterns in performance reviews. Return dashboards to managers with coaching on evidence-based feedback.

Metrics that matter

• Promotion speed: How long people spend at each level before moving up (entry → manager → director).
• Stage conversion: The share of employees who advance at each step; highlight where progress is blocked.
• First-year impact: After a promotion, what percentage meet or exceed expectations in year one.
• Time to full productivity: How long a newly promoted leader takes to hit steady performance—and what the empty seat costs you until then.
• Audit gap: The difference between performance and “potential” scores by gender: your early warning that ratings, not talent, are the problem.

 Why it pays to fix this

The confidence premium feels intuitive as charismatic leaders do inspire trust, after all. But intuition is expensive when it systematically mislabels your highest performers as lower potential. The fix isn’t to coach women to project more confidence. It’s to stop treating confidence as evidence of capability. This isn’t diversity theater. It’s pipeline quality control. Companies that convert performance into advancement using clean, auditable criteria build deeper leadership benches. They experience fewer flameouts among newly promoted managers. They shorten the time-to-impact on critical work. In markets that reward disciplined execution, competitive advantage may come down to something deceptively simple: raising the evidentiary bar for everyone.

For executives and boards, the question isn’t whether this bias exists. The data has settled that. The question is whether you’re willing to measure what you currently take on faith, and act on what you find.

Professor Ginka Toegel is Professor of Organisational Behaviour and Leadership at IMD Business School and author of The Confidence Myth: How Women Leaders Can Break Free from Gendered Perceptions (Palgrave Macmillan)