Women in leadership: one is never enough

For most of my twenty-five years in this business I have been the only woman at the table. Not one of a few. The one. I have sat in boardrooms where the ratio was me and eleven men, and I have listened to a room full of people explain the female consumer to each other with total confidence and no data, and I have watched work aimed squarely at women get approved by a group of people who had never once been her.

I used to think the fix was getting the seat. Then I got the seat and learned the more uncomfortable thing. A seat is not a voice, and one voice is not a majority.

There is research on exactly this, and it is far more specific than the version that shows up in diversity decks. It does not say that women on boards are generally beneficial. It says something much sharper and much more useful, which is that one of us, alone, changes almost nothing at all.

Twenty-eight days

A team of researchers from Lehigh, Notre Dame, Indiana and Auburn looked at 4,271 medical product recalls across 92 FDA-regulated public companies between 2002 and 2013. They wanted to know whether the composition of a board changed how fast a company pulled a dangerous product off the market. Not how it talked about safety. How fast it acted.

Firms with women on the board announced recalls of the most serious, life-threatening defects 28 days sooner than all-male boards, a 35 percent reduction in the time between knowing and deciding. In a product category where the defect can kill someone, twenty-eight days is not a governance statistic. It is a body count.

Here is the finding that should be on a poster in every boardroom in America. When a board had exactly one woman, seriously defective products were not pulled any faster. The effect did not appear. At two women it appeared. At three it was larger.

Two caveats I am going to give you rather than let you find them. This is an observational study, so it establishes an association and not a cause. Boards that appoint a second woman may differ in other ways that also make them decisive. And the threshold is specific to the most severe recalls: for lower-severity, more discretionary recalls, even one woman was associated with substantially more recalls being announced. So the honest headline is narrower than the one I want to write, and it is still the most useful thing on the subject I have ever read.

I want to sit with that, because it is the single most clarifying thing I have ever read about my own career. Every company that appointed one woman and issued a press release about it was, on this measure, buying nothing. The lone woman is not a diversity strategy. She is a rounding error with a nameplate, and I have been her more times than I would like to count.

Why one does not work

If you have never been the only one in a room, the threshold effect can look like a statistical curiosity. It is not. It is the most predictable social physics I know, and it operates on anyone who is the only anything.

When you are the only woman at the table, disagreement reads as a personality trait. You are not raising a concern about the campaign. You are being difficult about the campaign. The objection gets attached to you rather than to the work, and after it happens twice you start doing the arithmetic before you speak: is this hill worth the reputation, do I have the capital this quarter, will this be the thing that gets me described as hard to work with in a performance review written by someone who has never once had to make this calculation.

The second thing is subtler and worse. As the only one, you get read as a representative rather than an expert. Your view is not treated as professional judgment built over decades. It is treated as testimony from your demographic. "What do women think of this?" is a question that sounds inclusive and functions as an outsourcing of research to an employee, and it quietly reclassifies your expertise as anecdote.

Add a second woman and both mechanisms break. Now a disagreement is a position rather than a mood. Now two people can share the cost of saying no. Now nobody can be dismissed as the one who always brings this up, because there are two of you and you do not agree with each other about everything, which is precisely the proof that you were never speaking for a category in the first place.

This is why headcount is the wrong metric and I would rather you never used it. The question is not how many women are in the room. The question is how many people in the room can kill an idea and survive it.

Meanwhile, the room is emptying

You would expect this to be a story about slow progress. In 2026 it is a story about reversal.

In the first quarter of 2026, women's share of Russell 3000 board seats fell back below 30 percent, the first time it has slipped under that line since reaching it. The proportion of new Russell 3000 board seats going to women fell from 42 percent to 33 percent between 2022 and 2025. Disclosure of board gender data among S&P 500 companies collapsed from 91 percent to 60 percent in a single year, which means roughly a third of the largest companies in the country decided the safest move was to stop telling you. And in 2026 so far, white men have taken about 60 percent of new S&P 500 directorships, the largest share in a decade.

The institutional scaffolding came down too. BlackRock abandoned its 30 percent target for diverse directors. ISS, whose recommendations move an enormous share of institutional votes, suspended diversity as a factor in how it advises investors to vote on directors. When the machinery stops asking, boards stop answering.

And 42 percent of Russell 3000 companies have two or fewer women on the board. Hold that against the research and the picture is stark: two in five American public companies are sitting at or below the exact line where the benefit starts.

The number people quote as progress is the Fortune 500 CEO count. A record 55 women now run Fortune 500 companies, about 11 percent, the highest in the list's 72-year history. It is a real milestone and I do not want to diminish the women who earned it. I would also point out that a record is being set at 11 percent, which means the same headline reads: 89 percent of the largest companies in America are still run by men, and we are calling it a record.

The customer you are guessing at

Now the commercial part, because this is a branding journal and I am not here to make a moral argument that a CFO can decline.

In 2025, according to NielsenIQ and World Data Lab, women crossed half of all global consumer spending for the first time in recorded history. NIQ puts annual spending controlled by women at roughly 31.8 trillion dollars, and estimates that women make or influence 70 to 80 percent of purchasing decisions depending on the category. Reasonable people argue about the precise denominator. Nobody argues about the direction. There is no larger customer in the world economy, and there is no second place.

Brands still staff for her the way you would staff for a niche.

I have sat through the meeting where a room of men debated whether a woman would find a product "empowering." I have watched a campaign concept survive because nobody present had the standing to say that it read as condescending, and I have watched that same campaign land exactly as badly as the one person who flagged it said it would. Bic learned this publicly in 2015, when its South African account marked National Women's Day with an ad instructing women to "look like a girl, act like a lady, think like a man, work like a boss." The company defended it, then deleted it, then apologized. Somebody wrote that line. Somebody approved it. Somebody scheduled it. At no point did the process contain a person with both the perspective and the authority to stop it.

The receipts

Tone-deaf campaigns are the visible version of this problem. They are also the least expensive version, because a campaign can be pulled. The costly failures happen further upstream, in the product itself, and they are much harder to undo.

I want to be precise about what these prove and what they do not. I cannot show you a document from any of these companies confirming that a homogeneous team caused the omission. Companies do not publish the roster of who was in the room, and any writer who tells you otherwise is inferring. What I can show you is a pattern with no competing explanation on offer: over and over, products got built, tested, approved and shipped with a gap in them that is obvious the moment somebody who lives in a female body looks at it. The blind spot is not a mystery. It is a staffing outcome.

And notice that not one of these is a marketing problem. They became marketing problems. They started as design problems, which started as decision problems, which started as who was standing in the room.

What it does to your talent

The other cost is the one that shows up in your recruiting funnel eighteen months later, quietly, with no line item.

People look at your leadership page before they take your call. Not out of idealism. Out of forecasting. A woman evaluating your company is trying to work out whether there is a version of her future that ends anywhere near the top of your org chart, and the photograph of your executive team answers that question faster and more honestly than any careers page ever written.

The pipeline problem starts long before anyone is board-eligible. Eleven years of data from McKinsey and LeanIn shows the break is at the very first promotion to manager: 93 women promoted for every 100 men. At entry level, 31 percent of women have a sponsor against 45 percent of men, and sponsored employees are nearly twice as likely to be promoted. Only about half of companies still say they are prioritizing women's advancement at all.

The 2025 report found something that will get misused, so let me get to it first. For the first time, women reported lower interest in promotion than men. That will be quoted at you as evidence that women are opting out. Read the next line of the study: the gap falls away when women receive the same career support men do. Eighty percent of women want the promotion against 86 percent of men, and that six point difference tracks the support they are given rather than the ambition they arrived with. It is a sponsorship gap wearing an ambition costume.

The part where I am honest about the business case

You have seen the statistic. Companies in the top quartile for executive gender diversity are 39 percent more likely to financially outperform. It is McKinsey's, from the fourth edition of Diversity Matters, and it gets cited in roughly every deck on this subject including ones I have sat through.

I am not going to lead with it, and you should know why.

In 2024, Jeremiah Green and John Hand published a replication attempt in Econ Journal Watch. Using S&P 500 data, they could not reproduce McKinsey's findings. To be precise, because precision is the whole point of this section: they tested the 2015, 2018 and 2020 editions rather than the 2023 one, and they looked at racial and ethnic diversity rather than gender. McKinsey declined to share its underlying data or name the firms in its sample, and responded that its work shows correlation rather than causation. That response is fair. It is also a much weaker claim than the one being made in the decks. Similar correlational studies from consulting firms carry the same structural problem: variables are often self-reported, the direction of causation is rarely addressed, and the firm publishing the research also sells the remedy.

So I will not lead with a number I cannot defend, because the moment a skeptical CFO pulls that thread, everything else I said comes with it. And I will hold my own evidence to the same standard. The recall study is peer-reviewed, specific, and found a threshold, but it is observational too, and I am not going to pretend it proves causation while criticizing someone else for exactly that.

Here is what survives the standard. The product failures are not correlations. They are records. The Ambien dose is a matter of FDA record. The crash dummy timeline is in NHTSA's own report to Congress. The classification error rates are published and reproducible. The drug withdrawals are in a GAO report with a document number. You do not need a contested profitability statistic when you have a drug that was dosed wrong for half its users for twenty-one years.

Leadership dictates culture. Culture is the brand.

Everything above collapses into one idea, and it is the thing I actually believe after twenty-five years of building brands for other people.

A brand is not a logo, a palette, or a campaign. It is a promise an organization makes and then either keeps or breaks, thousands of times a day, in decisions no customer ever sees. Who gets hired. Which complaint gets escalated. Which product ships with the rough edge still on it. Which concern gets raised in the meeting and which one gets swallowed because raising it costs more than it is worth.

Those decisions are made by the culture. The culture is set by the leadership, not by the values printed on the wall. And the brand is simply what all of that looks like from the outside, over time, to a stranger.

Which means you cannot build a brand for women out of an organization that does not have any in the room where things get decided. You can build a campaign that way. You can buy media and hire a photographer and license a font. But the promise will not hold, because there is nobody inside the building whose lived experience makes the promise true, and customers work that out faster than any of us are comfortable admitting.

What I would tell my younger self

I spent a lot of years believing that if I were excellent enough, the ratio would stop mattering. That if I prepared harder, brought better numbers, and was more right than everyone else in the room, the room would eventually reorganize itself around the quality of the argument.

The room does not do that. The room does what rooms do.

What I would tell her is this. Being the only one is not a distinction, it is a structural problem, and no amount of individual excellence solves a structural problem. The work is not to be the best woman in the room. The work is to change the number.

So if you are a founder reading this with a board to build, or a CEO with a slate to fill, or a woman doing the arithmetic before she speaks in a meeting tomorrow morning: the number is two. Two is where it starts working. Three is where it holds.

And if you are sitting on a board right now looking around and counting, and the answer you get is one, you already know what I am going to say. You do not have a woman on your board. You have a witness.

Count the room. Then change the number.
Brea