How to Find Academic Conference Speakers

The most common failure in speaker selection is inviting good people and discovering afterwards that the programme has a hole in it. Write the shape of the event first, then search against the gaps, exactly as a grant work plan is written before the partner list.

Write each session as a question the audience should be able to answer afterwards, not as a topic label. A question is searchable; a label such as advances in the field is not.

Decide the format per slot before the person

a plenary that frames the field, a session that reports new results, a methods tutorial, a panel that stages disagreement. The format changes who is a good choice far more than seniority does.

Note which perspectives the event needs and does not yet have: a clinical view alongside a technical one, an industrial or policy view, a critic of the dominant approach, a neighbouring discipline that solved the same problem.

Set the numbers early. How many invited slots, how many come from the open call, how many are held for late-breaking work. Committees that leave this vague end up with a programme dominated by whoever answered first.

Write the constraints down as well

budget per speaker, whether travel and accommodation are covered, whether remote presentation is possible. These decide more acceptances than prestige does, and they belong in the first invitation rather than in the third exchange.

Four sources between them produce nearly every good name that is not already in the committee's head. None of them is expensive; all of them take an afternoon rather than a meeting.

The recent literature on the session question, read as a list of people rather than of papers. The last two or three years matter more than career totals, because you are looking for who is doing this work now, not who defined the field a decade ago.

Programme committees and speaker lists of past editions, and of adjacent conferences in neighbouring fields. Reading who other events invited is the fastest way to find names your community has not used, and reading your own past programmes tells you who is being repeated.

The people around the senior names. Track the first authors of the papers you admire, the postdocs and junior faculty presenting the actual work, and the people whose posters drew a crowd at the last edition. A first author who has published three papers on your session question is often the better speaker on it.

Communities that sit outside your usual reading

a methods community whose tools your field uses, a regional or non-English-language community working on the same problem, practitioners in a clinic, an agency or a company. These names never surface from a keyword search in your own vocabulary.

Our conference speaker finder serves the first of those routes. Upload the call for papers, a track description or a paper that defines the theme, and it suggests researchers from the public literature whose work sits close to it, with a short reason each. It is a widening tool and nothing else: it says nothing about availability, willingness, speaking ability or whether someone has spoken at your event before, and it contacts nobody.

A long list is not a programme. Balance is a set of deliberate trade-offs made across the whole event, and a committee that evaluates candidates one at a time will produce an unbalanced result no matter how good each individual choice is.

Lay the whole draft programme out at once and look for duplication. Two speakers who would give the same talk is the most common structural fault, and it is invisible when invitations are decided one by one.

Balance career stages on purpose. A programme entirely of senior figures is predictable, and one entirely of early career researchers loses the framing talks that make the rest legible. Mixing stages within a session works better than segregating juniors into their own track.

Balance institutions and countries so that a single group does not hold three slots, which is the usual result of a committee inviting from its own networks.

Balance the intellectual position. A programme where everyone agrees is comfortable and forgettable. Inviting a well-argued critic of the dominant approach is the single cheapest way to make an event worth attending.

Keep a ranked reserve list per slot before the first invitation goes out. Declines are normal, they arrive weeks later, and a committee without a reserve list falls back on whoever it can reach, which is how programmes drift towards the usual names.

Sequence the invitations. Send the ones that anchor the programme first, because a confirmed keynote makes every subsequent invitation easier to accept.

Speaker imbalance is well documented and it is structural rather than deliberate. Two findings are worth knowing because they point at fixes rather than at blame.

Composition of the committee affects the outcome. Casadevall and Handelsman found that symposia with at least one woman among the conveners had a significantly higher proportion of women speakers than those convened by men alone, across a large sample of microbiology meetings (mBio, 2014).

The gap is not explained by the available pool. Nittrouer and colleagues found that women were underrepresented among colloquium speakers at leading United States universities relative to the pool of eligible faculty, and that the difference was not accounted for by women declining invitations more often (PNAS, 2018).

The practical implication is procedural

diversify the committee, build a long list before deciding rather than after, and count the draft programme on the axes you care about while you can still change it.

Ask every invited speaker for one name they would recommend on the same topic. It is the cheapest single intervention available and it consistently returns people the committee had not considered.

Directories exist for the recurring objection that nobody could be found. Request a Woman Scientist is one, and many societies and communities maintain field-specific lists.

Remove the barriers that produce declines rather than treating declines as preference: state travel support up front, offer remote or pre-recorded formats, publish dates early enough to be planned around, and where you can, provide care support or accessible scheduling.

A publication record answers one question well and several others not at all. Ten minutes of checking per candidate prevents most of the awkward outcomes, and none of it requires contacting the person.

Check that they still work on the topic. A highly cited paper from six years ago is not evidence of current activity, and inviting someone to speak on work they have moved on from produces a flat talk.

Watch or listen to something. Recorded talks, seminar videos and conference recordings are widely available, and fifteen minutes of one is the only real evidence of speaking ability that exists before you commit a plenary slot.

Check your own history. Committees regularly re-invite someone who spoke two editions ago because nobody kept the list. Keep a record of past speakers with dates and topics.

Check for the obvious conflicts and optics

a speaker with an undisclosed commercial stake in the topic, or a session where every speaker is from the same collaboration.

Check the practical constraints before promising anything

whether your budget covers their travel, whether visa timelines are realistic for their country of residence, and how far ahead they need to plan.

Where the event carries reputational risk, do the basic public checks you would do for any invited guest. Nothing here is done by a matching tool, and every one of these judgements belongs to the committee.