The Data Behind Why Women Get Invited to Speak Less Often
When an all-male panel gets called out online, the organiser's defence is almost always the same. We reached out to women, they said no, we did our best. It sounds reasonable. It is also, according to the research on how speaker lists actually get built, only half the story, and often not the more important half.
The more interesting question isn't why women decline. It's who gets asked in the first place, and by whom.
The Default Is Not Neutral
Every panel starts with someone picking up their phone, or opening their contacts, or scrolling LinkedIn, and thinking of names. That process feels like a search for the most qualified person available. In practice, it runs on proximity. Organisers reach for people they already know, people who were recommended by someone they know, or people who showed up at the last three events they also attended. None of that is a competence filter. It is a network filter, and professional networks in most fields are still shaped by who has historically held the senior, visible roles.
Researchers have a name for this: homophily, the tendency of people to connect with and recommend others who resemble them. It isn't unique to gender. People cluster by seniority, by institution, by geography, by industry vertical. But when the senior, visible layer of a field skews male, and the informal recommendation chain runs through that layer, the "obvious first names" for a panel skew the same way, round after round, event after event. Nobody has to be deliberately excluding anyone for this to produce a stage full of men. The system does that on its own, by default, unless something interrupts it.
That's the part the "we asked and they said no" defence conveniently skips. It answers for the invitations that were sent. It says nothing about the ones that never got considered in the first place.
What Happens When the Person Building the List Is a Woman
If the pipeline explanation were the whole story, the gender of the person building the speaker list shouldn't matter much. The research says otherwise, consistently, across very different fields.
A long-running study of two North American scientific societies, the Society for Conservation Biology and the American Society of Ichthyologists and Herpetologists, tracked symposium organisers and symposium speakers over more than a decade. It found that women's share of speaking slots tracked closely with women's share of the organising committee, roughly a third of both at one society and closer to a fifth at the other. When the study looked at newer medical-conference research across multiple specialties and regions, it found a significant positive correlation between the proportion of women on an event's planning committee and the proportion of women who ended up speaking, alongside a broader finding that more than a third of panels in that dataset were composed entirely of men.
Oncology offers one of the more granular pictures of what this looks like inside a single field. A review of four consecutive years of programme data at a major international oncology conference found that of 670 sessions, 81, or just over 12 percent, were all-male. That is a field with no shortage of accomplished women publishing research, treating patients, and leading trials. The manels in that dataset weren't a symptom of a missing pipeline. They were a symptom of who kept getting asked back, session after session, year after year.
None of this means male organisers are acting in bad faith, or that women organisers are somehow immune to the same defaults. It means the search itself is the variable that moves the outcome, not the underlying pool of qualified people. Widen who is doing the searching, and the list that comes back looks different, because a different set of names sits close enough to reach.
Where This Shows Up in India
India's version of this gap does not look like a shortage of accomplished women. It looks like a gap between ambition and access, opening up at the exact point where visibility starts to compound. The second edition of the Women Leadership in Corporate India Survey, run jointly by AIMA and KPMG in India in 2026, found that 79 percent of women professionals surveyed said they aspire to hold a leadership role, while only 1 percent currently hold a board-level position. That is not a pipeline problem in the sense of too few women entering the workforce. It is a filtering problem, happening somewhere between aspiration and the rooms where board seats, keynote slots, and speaking invitations actually get decided.
Conference stages sit downstream of that same filtering. A woman who hasn't been made visible through board seats, bylines, or repeated media appearances is less likely to be the "obvious first name" an organiser's network throws up, even when her actual expertise matches the brief precisely. The two gaps reinforce each other. Fewer women in the rooms where recommendations get made means fewer women recommended, which means fewer women on stage, which means fewer women visible enough to be recommended the next time a similar brief comes around.
This is the argument this platform keeps returning to: the shortfall in who gets invited to speak is rarely a shortfall in who exists. It's a shortfall in who the search reaches, and that gap runs through corporate India's leadership numbers as clearly as it runs through any single conference programme. We've written before about how institutional headcount gains for women don't automatically translate into podium time, and the organiser-network data is a large part of the mechanism behind why that translation keeps failing. The same pattern shows up outside the conference circuit too. In sectors like AI-driven healthcare, where women are building the actual systems, the technical work and the public credit for it have a way of drifting apart, for exactly the same reason.
Declining Isn't the Whole Explanation Either
It's worth taking the "women say no" defence seriously enough to examine it properly, because there is some truth buried in it, just not the whole truth. The conservation-biology symposium research found that women were somewhat more likely than men to decline a speaking invitation once one was extended, and that men were, in some circumstances, less likely to accept an invitation that came from a woman organiser rather than a man. Both patterns are real, worth investigating in their own right, and connected to everything from caregiving load to the discomfort of being the only woman on a panel, to the plainer fact that being under-recognised for years makes some women hesitate to call themselves an expert out loud, even when they clearly are one.
But decline rates only explain the gap between invited and confirmed. They explain nothing about the much larger gap between exists and invited in the first place, which is where homophily does its quiet, compounding work. An organiser who only reaches three names deep into a shallow, familiar network will keep landing on the same three men, regardless of how many women would have said yes to a fourth or fifth invitation that never got sent to them.
What a Wider Search Actually Requires
None of the research suggests organisers need to lower the bar for who qualifies to speak. It suggests they need to widen the search radius before the bar gets applied at all. A field can be full of qualified women and still produce all-male panels indefinitely, if every list is built from the same six phone numbers, recycled event after event.
Practically, that means treating speaker sourcing the way any other sourcing decision gets treated, as a process that needs deliberate inputs, not just instinct and familiarity. It means going past the first three names a personal network produces on the first pass. It means building and maintaining lists that don't depend entirely on any one organiser's personal contacts, so the search doesn't reset to zero every time a new person is doing the inviting. Treating visibility infrastructure as something that has to be built on purpose, because the data makes clear it does not build itself, is the whole premise behind why a searchable directory of vetted women experts is useful in the first place.
That kind of directory shortens the search radius, so the "obvious first name" an organiser reaches for isn't limited to whoever happens to be three degrees away from them on LinkedIn. It doesn't fix the deeper structural gap in board seats and bylines on its own, nobody has built a tool that does that in one step. But it removes one very specific, very fixable excuse: that the organiser looked, genuinely, and simply couldn't find anyone who fit.
There's also a compounding benefit worth naming directly. Every time a wider search surfaces a woman who wasn't on anyone's shortlist before, she becomes part of the next organiser's shortlist too, whether that organiser finds her through a directory, a shared panel, or a colleague who saw her speak. Visibility infrastructure works the same way networks do, just in the opposite direction. Homophily narrows the pool by default. A deliberately maintained, searchable pool widens it the same way, on purpose, one appearance at a time. That is a slow mechanism, not a switch that flips overnight, but it is the same mechanism that produced the narrow version of the problem in the first place, running forward instead of backward.
The Honest Version of the Excuse
The organisers who say "we tried and couldn't find women" are, in most cases, telling the truth about their own search. They asked the names that came to mind first, and those names ran out quickly. What the research makes clear is that this isn't a statement about who exists in a field. It's a statement about how far a familiar network reaches, and whose names sit inside it by default, long before anyone gets to the point of extending or declining an invitation.
That's a fixable problem, but only if it gets named correctly. Treating it as a talent shortage leads organisers to run the same search again and arrive at the same result. Treating it as a discovery problem, caused by who is doing the looking and how far they're willing to look beyond their own contacts, is the version of the excuse that actually points somewhere useful. The data on organiser composition backs that up again and again, across scientific societies, medical conferences, and oncology programmes that have nothing else in common with each other. Change who's searching, or how widely they search, and the panel changes with it.
Draupadi on the Dais connects accomplished Indian women experts with the panels, media, and platforms built to find them. Claim the Dais. Follow us at @draupadionthedais.