Somewhere this week, a woman who is very good at her job will stand in front of a mirror and recite her accomplishments out loud, because an article told her that is how you treat impostor syndrome.
She will go back to a performance review that called her lacking in leadership presence and offered no example. She will go back to a calibration meeting she was never in, where a decision about her was already made. She will go back to a team where she is the only one of her kind in the room and has been for four years.
The mirror will not touch any of that.
What the 1978 Paper Actually Said
Pauline Rose Clance and Suzanne Imes published their study in 1978. They worked with high-achieving women who, despite outstanding academic and professional records, persisted in believing that they were really not bright and had fooled anyone who thought otherwise. They called it the impostor phenomenon. The word syndrome came later, from everyone else.
It was a careful clinical observation of a specific group, and it described something real. Women who had the evidence in hand and could not hold it.
What the study did not have was the rest of the world. As Ruchika Malhotra and Jodi-Ann Burey put it in the piece that started this argument, "the impact of systemic racism, classism, xenophobia, and other biases was categorically absent when the concept of impostor syndrome was developed." The sample left out women of colour and women across a range of incomes, genders and professional backgrounds. A description drawn from a narrow room got applied to every room there is.
And then something happened to it that happens to good ideas. It escaped.
Forty-seven years later it is an industry. Search for it and you will find millions of results, offering conferences, books, workshops, coaching packages and the mirror. Famous women have confessed to it — actors, chief executives, a former First Lady, a sitting Supreme Court justice. Rarely is a woman invited to a career development conference where a session on overcoming impostor syndrome is absent from the agenda.
What almost nobody sells is an explanation of why so many women feel this way in the first place, and what the building they work in has to do with it.
A description became a diagnosis. A diagnosis locates the fault inside the patient. And once the fault is inside the patient, every available cure is sold to her.
The Two Words
The phrase itself does work that nobody audits.
Impostor. The word carries criminal fraudulence. It belongs to a person who has assumed a false identity in order to obtain something they have no right to. That is a heavy coat to put on the ordinary feeling of being new to a team, or learning a skill you have not learned yet, or sitting in your first meeting two levels above where you sat last year.
Syndrome. A cluster of symptoms. Something you suffer from. Malhotra and Burey make the sharpest observation in the piece here: the medical undertone recalls the female hysteria diagnoses of the nineteenth century, a label that converted inconvenient women into patients for the better part of a hundred years.
Put the two words together and you have taken a universal professional experience — uncertainty, second-guessing, mild anxiety in a new room — and pathologised it, mostly for one half of the workforce.
Here is the asymmetry that makes it indefensible. A man's doubt tends to resolve on its own, because the environment resolves it for him. His work is validated over time. He can find role models who look like him at every level above him. His competence, his contributions and his leadership style are rarely questioned. The same doubt, in the same industry, on the same timeline, receives the opposite treatment: it is questioned, it is tested again, it is fed back to him never and to her constantly, and eventually she is told she has a condition.
Doubt that the environment refuses to resolve stops being a personality trait and becomes a reading of the environment.
If the same person feels like an impostor on one team and competent on the next, with the same skills and the same résumé, the variable is the team. This is the single most useful diagnostic in the whole subject, and it takes thirty seconds to run.
The Numbers Went the Other Way This Year
Any argument about whether a problem is individual or structural can be settled by counting.
The headline from the United States is a record. Fifty-five women now run companies on the Fortune 500 list — 11 percent, the highest share in the list's seventy-two-year history, and the fourth consecutive year above ten percent.
The same sentence, read from the other side of the ledger: after seventy-two years, the largest companies in the world's largest economy are run by men in about nine cases out of ten. On the Fortune Global 500 the figure is 33 women, or 6.6 percent.
And underneath the record, the direction of travel has reversed.
| Measure | Latest | Year before |
|---|---|---|
| Russell 3000 board seats held by women | 29.2% | 30.1% |
| Russell 3000 new board appointments going to women | 22.7% | 29.1% |
| S&P 500 board seats held by women | 34.4% | 34.6% |
| S&P 500 new directors who are women | 29% | 38% |
That first line is the story. Women's share of board seats across the 3,000 largest American public companies has declined for the first time in more than a decade.
The second and fourth lines are the forecast. New appointments are the flow that feeds the stock, and the flow has fallen off a cliff — from 29.1 percent to 22.7 percent in the Russell 3000, and from 38 percent to 29 percent among incoming S&P 500 directors. Board seats turn over slowly, so a collapse in new appointments shows up in the totals for years afterwards. The decline you can see today is the small part. The decline already committed is larger.
Thirteen years of progress bought roughly thirty percent of the seats, and one year of retreat is taking it back faster than it was won.
No individual woman's confidence moved those numbers. No mirror will move them back.
The Rung Before the Ceiling
The boardroom is the last door. The one that matters more is the first.
McKinsey and LeanIn have run their Women in the Workplace study for eleven years, and their central finding has a name: the broken rung. It sits at the very first promotion into management, long before anyone is near a glass ceiling.
In the 2025 edition, for every 100 men promoted into a manager role, 93 women were. For women of colour the figure is 74. Those gaps compound, because a promotion missed at the first rung is not recoverable later — it removes you from the pool that the second rung draws from. Run 93-per-100 for a decade and the pipeline does not need a ceiling to produce a male executive team. Arithmetic is sufficient.
Sponsorship runs the same way. At entry level, 31 percent of women have a sponsor, against 45 percent of men. The distinction matters more than almost anything else in a career and is almost never taught: a mentor talks to you, a sponsor talks about you in rooms you are not in. Advice is cheap and widely distributed. Advocacy is scarce and distributed to people who remind the advocate of themselves.
And then there is the finding from the 2025 report that will be misread more than any other. For the first time, women are notably less likely than men to say they want to be promoted.
That fact will not survive contact with the industry intact. It will be called an ambition gap. It will be called a confidence problem. There will be sessions about it.
There is another reading. A person who has watched the first rung pay out at 93 cents on the dollar, who has a sponsor at a rate of 31 percent against someone else's 45, who has received feedback with no example attached and seen a peer promoted on a presentation style she is punished for using — that person declining to invest more effort in the same machine is pricing it correctly. When a system reliably underpays a particular kind of effort, reducing that effort is rational.
Which is the whole argument in one move. Call it a confidence problem and you will buy a workshop. Call it an accurate assessment of returns and you will have to fix the returns.
The Shape of It in Technology
For anyone reading this from a tech lead or engineering manager's chair, the shape is sharper and the drop is earlier.
Women hold somewhere around 26 to 28 percent of technology roles to begin with. Across industries the staircase runs roughly 51 percent of entry-level roles, 39 percent of manager roles, 29 percent of executive roles. In engineering specifically, recent work describes a director ceiling: women make up about 14 percent at manager level and about 9 percent at director — the step where a career stops being about building things and starts being about controlling budget, headcount and roadmap.
That step is where organisational power actually begins, and it is the step with the steepest cliff.
There is a mechanism behind it that most engineering organisations have never examined. Somebody has to do the work that holds a team together: the onboarding, the design review nobody volunteered for, the incident write-up, the cross-team negotiation, the documentation, the mentoring. Tanya Reilly named it glue work, and her observation is the one that matters here. Glue work is indispensable, it is praised warmly in the moment, and it is almost never what a promotion packet is written on. Promotion packets are written on systems shipped and scope owned.
The question worth asking is who gets handed the glue, who is thanked for it, and who is told two years later that their impact is hard to point to.
An organisation that distributes glue work unevenly and then promotes on artifacts has built a filter and called it a meritocracy. The filter runs without anyone intending it, which is exactly why it survives every diversity initiative that does not look at it.
India: A Mandate Met Exactly as Written
Now turn to India, where the story is worse, and worse in a way that teaches something.
In 2013, around 6 percent of board seats across the NIFTY 500 were held by women, and roughly 60 percent of those boards had no woman on them at all. Then the Companies Act of 2013 required certain classes of company to appoint at least one woman director, and the market regulator later extended it so that the top listed companies needed at least one independent woman director.
A decade on, women hold about 21 percent of NIFTY 500 board seats. Roughly tripled. A regulatory success, if you stop reading there.
The layer underneath changes the picture.
| Indian boardrooms | Figure |
|---|---|
| Women's share of NIFTY 500 board seats, 2025 | ~21% |
| …of whom are independent rather than executive directors | ~60% |
| Women directors in the BSE 200 who are executive directors | ~11% |
| BSE 200 chairpersons who are women | ~7% |
| Women CEOs in NSE 500 companies | just over 7% |
An independent director attends a handful of board meetings a year, chairs a committee or two, and holds no operating authority over anything. An executive director runs a business. The law asked for a woman on the board and said nothing about which kind, so corporate India supplied the cheapest compliant answer available: a seat, filled correctly, attached to no power.
One of the better recent studies of this calls the gap exactly what it is in its title — the distance from presence to influence.
A mandate written as a headcount will be met as a headcount. If you want power shared, you have to write the rule about power.
This is a lesson that travels far beyond India and far beyond gender, and anyone who has ever watched a compliance requirement land in an engineering organisation already knows it. Specify the artifact and you get the artifact. Specify the outcome and you might get the outcome. Representation targets measured as headcount produce headcount — a hire made, a seat filled, a number reported in an annual report — while the thing everyone claimed to want, which is decisions being made by a wider set of people, goes unmeasured and therefore unbuilt.
If You Are the One Being Diagnosed
Everything above is structural, and you did not build any of it. But you still have to go to work tomorrow, and a structural explanation that leaves you with nothing to do on Monday is its own kind of failure. So here is the practical half, written for the engineer who is a lead or a manager now and intends to go further.
Keep the record, and keep it written. A ledger rather than a diary. What shipped, when, what it was worth, who it unblocked, what broke and what you did about it. Do this monthly, because reconstructing a year in the week before a review is how good work disappears. The purpose is specific: vague feedback is unfalsifiable, and a written record forces a conversation to become specific or collapse.
Convert unfalsifiable feedback into observable behaviour. When you are told you lack leadership presence, or executive presence, or that you are not quite ready, ask three questions, calmly and in writing afterwards. What were the specific instances? What would the alternative behaviour have been in that moment? What would have to be true for the answer to be yes? Feedback that can survive those three questions is a gift and you should take it seriously. Feedback that evaporates under them was never about you, and you now have that in writing.
Understand that the decision happens before the meeting you are in. Promotions in most large organisations are settled in calibration, among managers, with limited currency and competing candidates. Your manager walks into that room needing two things: ammunition and an ally. Your self-belief is in neither category. Give them a one-page case they can read out — scope owned, outcomes delivered, the next level's behaviours already demonstrated — and you have changed what they can say about you when you are not there.
Ask for sponsorship by name, and make it easy to say yes. Few things are more uncomfortable and few things pay better. Will you mentor me costs them a coffee and buys you advice. The version that works sounds like this: I am going for staff engineer next cycle. Would you be willing to speak to my work on the migration in calibration? It is specific, it is bounded, it names the room, and it lets someone agree to a concrete act rather than an open relationship.
Watch the glue. Do some of it — a team that nobody holds together is a worse place to work and you will not want to be the person who refused. But keep a ratio, know what it is, and make it visible rather than invisible. Write the onboarding guide as a documented deliverable with your name on it. Run the incident review and publish the findings. Glue work that leaves an artifact counts twice; glue work that leaves nothing counts once, in the moment, in a conversation nobody remembers.
Choose problems that survive a change of manager. Reorganisations happen. The manager who knew what you did leaves. What remains is the system that is still running, the revenue attributable to it, the incident class that stopped occurring. Build those, and your case travels with you.
And treat the doubt as instrumentation. Doubt about a specific thing you have not done before is information, and it is usually correct and useful. Doubt that attaches to everything, in one particular room, among one particular set of people, while the same skills felt perfectly solid eighteen months ago on a different team — that is a reading of the room. Instruments are supposed to respond to their environment. Yours is working.
None of this is a substitute for the building being fixed. It is how you do well while the building is what it is, which is the situation nearly everyone reading this is actually in.
If You Run the Calibration Meeting
And if you are on the other side of the table — a manager, a director, anyone who sits in the room where these decisions are made — the work is different, smaller and more uncomfortable, because it is entirely within your control.
Stop funding sessions on overcoming impostor syndrome. They locate the defect in the person who has the least power to change it, and they are cheap in a way that is the point: a workshop costs a fraction of what fixing a promotion process costs, and it produces a photograph.
Measure the first rung, not the top of the house. The number almost every organisation reports is representation at senior levels, which moves slowly and tells you about hiring decisions made a decade ago. The number that tells you what your organisation is doing right now is the promotion rate into first-line management, split by group. If it is 93 per 100 in your company, you have found your problem and it is nowhere near the boardroom.
Refuse unfalsifiable feedback in writing. Make it a rule in your own calibration: no assessment goes into a packet without a specific observed instance and a statement of what the alternative behaviour would have been. Lacks executive presence fails that test instantly, every time, and it is the single most reliable vehicle for bias that performance management has ever produced. Banning it costs nothing and changes what people are able to say.
Audit the glue. Pull the last four quarters of work that held your teams together and look at whose name is on it. If the distribution has a pattern, you are running the filter described earlier. Then either rotate it deliberately or start counting it in packets, and do one of those two things rather than mentioning it warmly in a town hall.
Assign sponsorship rather than hoping for it. Sponsorship forms naturally between people who recognise themselves in each other, which is precisely why its distribution looks the way it does. It can be assigned. Name senior people, name the person whose work they will speak for, and ask them to report what they said in the room.
And remember what India's boardrooms demonstrated at national scale. A requirement written as a seat gets you a seat. If what you want is power shared, write the requirement about power — about who owns the budget, who owns the roadmap, who gets the assignment that generates the artifact that promotion packets are made of.
The Patient Was Never the Problem
Clance and Imes described something true. Women do carry this feeling, in enormous numbers, and describing it was a real contribution. The failure came later and belonged to everybody else: we took a description of a wound and filed it as a pre-existing condition.
Forty-seven years, millions of search results, a thousand conference sessions and an entire coaching economy later, the record on the thing those sessions were meant to change reads as follows. Eleven percent of the Fortune 500. Ninety-three women promoted into management for every hundred men. Board representation in the largest American companies falling for the first time in more than a decade. Twenty-one percent of Indian board seats, of which the overwhelming majority come with no operating authority at all.
If impostor syndrome were the cause, half a century of treating it would have moved one of those numbers.
The women in your organisation who second-guess themselves in meetings are reading an environment accurately and reporting it to you, in the only register the environment has left open to them. The instrument is fine. Go and look at what it is pointed at.
This is Part 1 of The Unwritten Rules, a series on how corporations actually work — power, promotion, and who gets to grow. It takes its starting point from "Stop Telling Women They Have Imposter Syndrome" by Ruchika Malhotra and Jodi-Ann Burey, published by Harvard Business Review in February 2021 and collected in their volume on women and leadership. The quoted passage is theirs; the argument, the data and any errors in either are mine.
