More women will lead—once they see what they’re missing

Every early-stage founder has heard it. Want more women in leadership? Run a confidence workshop. Launch a mentorship program. Push them to “lean in.” The logic sounds solid. The problem is execution. Most of these efforts don’t move the needle—at least not in the direction that matters: measurable, lasting behavioral change at the decision point. It’s not that women don’t want to lead. It’s that the system tells them—subtly, structurally, and sometimes strategically—not to bother.

A new research study led by Sophia Pink, a Wharton PhD candidate, just dismantled that coaching-heavy mindset. Her team found that women are more likely to step into competitive leadership opportunities not after being encouraged—but after being shown just how often they don’t. This single insight rips the heart out of a startup myth that’s wasted time, money, and talent for years. And it forces a hard truth on anyone running an org today: the bias isn’t in ambition. It’s in the design.

Pink’s study, published in Organization Science, wasn’t a sweeping policy change or a multi-stage coaching series. It was a sentence. A well-placed message that said:

“Women like you tend to compete less than men. That gives men an advantage.”

No slogans. No empowerment talk. Just data, context, and a moment of friction.

The result? In controlled tests, women who saw that message were significantly more likely to opt into competitive pay schemes—at the same rate as men. And on an executive job platform, women shown a similar nudge applied to 29% more top roles that day. What this reveals is executional gold: timing matters more than tone. And what you make visible shapes what people believe they can disrupt. Most leadership programs ask women to perform differently. This one asked the system to expose itself.

Let’s be clear: this isn’t about hating on coaching. It’s about understanding what coaching can’t fix. When startups build internal programs to “support” women into leadership, they often do three things:

1. Target the individual, not the environment
2. Focus on soft skills over decision friction
3. Treat outcomes as proof of progress—even if they’re isolated

These are all execution errors. Let’s walk through them.

When a system consistently under-represents women at the top, blaming individual readiness is lazy logic. Founders love “ownership” narratives. But you can’t self-help your way out of a rigged pipeline. Pink’s research shows that when women see how the odds are stacked, many push back—not because they’ve been encouraged, but because they’ve been underestimated. That’s not self-help. That’s systems defiance.

Too many leadership programs confuse confidence with capability. They try to close the gap by boosting self-perception. But perception alone doesn’t trigger behavior—especially when the system still favors someone else. Confidence training without structural visibility is just optimism with better lighting.

When one woman gets promoted after a coaching program, teams call it success. But success isn’t one hire. It’s whether the pattern changed. Did more women apply? Did more get through? Did more stay? Most don’t track that. They track what’s easiest to frame—and hardest to repeat. That’s startup theater. Not execution.

The idea that women need to be “fixed” before they can lead is a hangover from industrial-era org design. It frames women as the exception, not the system’s default. And every coaching effort built on that premise bakes fragility into the pipeline.

This shows up in subtle ways:

• Job descriptions that skew masculine-coded language
• Internal nomination processes that rely on visibility (and bias)
• Stretch assignments offered based on past confidence, not future potential

Each of these steps filters out women long before the final interview.

And here’s the kicker: most companies don’t see these filters. They see the result and try to solve it from the end, not the root. That’s like trying to optimize churn without fixing onboarding.

The success of Pink’s study didn’t come from better messaging. It came from timing and truth. Stereotype reactance—the mechanism at work—only activates when people are shown a stereotype they’re not supposed to fit. It flips the pressure. It creates an opportunity for defiance.

In startup terms: it inserts intentional friction at the choice point. You don’t need a 6-week curriculum. You need a sentence at the right moment. The best behavioral designs are invisible until they’re decisive. This is one of them.

Too many founders obsess over top-of-funnel numbers: “We interviewed 40% women last quarter.” Or “we have 5 women on the exec team now.” Those aren’t system metrics. They’re snapshots. And snapshots don’t show conversion, retention, or repeatability.

What matters is the conversion gap between equally qualified men and women when a role opens up. And what the internal feedback loop tells each group about their chances. Most systems whisper to men: “This is for you.” And to women: “You might not be ready yet.” What Pink’s study did was flip that whisper into a spotlight. And the system blinked.

You don’t need another initiative. You need a few specific rewires.

1. Put Reactance Messaging at the Moment of Decision

Whether it's a job board, an internal stretch assignment, or a promotion cycle—insert messaging that shows the pattern. Don’t sugarcoat it. Make the gap visible. Then ask: “Want to change that?” That’s it. No applause. No extra load on women. Just clarity.

2. Redesign for Choice, Not Compliance

Most DEI systems nudge toward participation. That’s fine. But participation under pressure isn’t empowerment—it’s performative compliance.

Design instead for intentional choice:

• Use opt-in structures with stereotype triggers
• Measure actual application or nomination behavior, not just workshop attendance
• Track who holds decision-making power—and who doesn’t feel invited

3. Stop Measuring Inputs. Measure Pattern Breaks.

The only question that matters: did this change the pattern? If your female promotion rate jumps after one nudge but falls back the next quarter, the system isn’t fixed. It’s just flinched. Track changes in repeated behavior—not just one-off lifts. And treat any success that doesn’t scale as suspect.

Investors love to tout inclusive teams—especially in LP letters. But few ask the right questions at portfolio check-ins. Here’s a better prompt: “What behavior shifted this quarter—and what messaging caused it?”

If the answer is a coaching program, dig deeper. If it’s a pattern-break in decision friction, you’ve got signal. What you want to see is system leverage, not surface optics. Boards should stop accepting DEI dashboards as proof. Ask for conversion audits and stereotype defiance metrics. They’re rare—but they exist. And they tell you who’s actually building a team for the long haul.

The most dangerous bias in startups isn’t the one that blocks access. It’s the one that pretends visibility equals equity. Having women in the room isn’t the win. Having them compete without self-doubt—or design disadvantage—is. Presence is a start. Pattern disruption is the goal. And systems don’t shift unless someone breaks them first.

Sophia Pink’s study didn’t uncover some magic formula. It just proved what smart operators already know:

People don’t need to be pushed into leadership. They need to be unshackled from the quiet lies that tell them they don’t belong there. For women, that lie has long been: “You’re not competitive enough.” Turns out, showing that lie for what it is does more than any motivational push.

The truth is the unlock. And choice—not coaching—is what scales. Most founders don’t need another program. They need to look at the system they built—and decide who it really favors when no one’s watching. If your system whispers “You don’t belong,” don’t build a louder invitation. Build a better room.