Speaking at a conference: getting selected and being good

A call for papers (CFP) is the process a conference uses to collect and select talk proposals, usually through a form asking for a title, an abstract and a short speaker biography. Most programme committees are working through anywhere from dozens to several hundred submissions against a fixed number of slots, so the proposal is doing a specific job: giving a reviewer, who has read thirty of these already that afternoon, a reason to keep yours.

What a reviewer is actually checking

A proposal clears the first pass on three things. Does the topic fit what this particular audience came for. Does the abstract describe a talk, with a shape and a conclusion. And does the speaker sound like someone reporting on work they have actually done. Reviewers are not scoring charisma at this stage. They cannot see you present, so the only evidence they have is what you write.

Writing an abstract that describes a talk

The most common weakness in a rejected proposal is a title and abstract that name an area. "Kubernetes at scale" is a subject. "Why our migration to Kubernetes took eighteen months longer than planned, and what we would tell a team starting today" is a talk, because it commits to a position and implies an ending. A working structure for the abstract itself is one sentence framing the problem the audience recognises, one sentence stating what you argue or found, and one or two sentences on what someone in the room will be able to do differently afterwards. Reviewers read the last part closely, because it is the difference between a talk and a topic dump.

Mistakes that get proposals rejected

• The vendor pitch in disguise. If the talk only makes sense with your product in it, reviewers can usually tell from the abstract, and most programmes exclude it on sight.
• The title that tries to cover everything. A talk that promises to explain "the future of X" in forty minutes reads as unfocused.
• No stated takeaway. An abstract that describes the journey but never says what the audience leaves with is the single easiest reason to pass on a proposal.
• One abstract sent everywhere unchanged. A proposal written for a technical single-track event rarely fits a mixed-audience conference, and reviewers notice when it has not been adjusted.
• Format mismatch. Pitching a workshop-length idea into a twenty-minute slot, or the reverse, signals that the speaker has not read the call for papers properly.

Telling a genuine CFP from a pay-to-speak slot

Some programmes marketed as open calls are pay-to-speak: the slot is sold as part of a sponsorship package, and the "selection" is really a sales conversation. There is nothing wrong with sponsored sessions as long as they are labelled as such in the programme, but a first-time speaker who submits to what looks like a competitive CFP and is instead asked to pay, or told selection is easier for sponsors, is in a different process than the one being advertised. Reading the CFP terms for any mention of sponsorship tiers or paid tracks before submitting saves the confusion later. Organisers who run their own call for papers face the same judgement calls from the other side, covered in Finding, Choosing and Looking After Speakers and Designing a Conference Programme.

A conference talk works when it carries one argument. Pick the single claim you want the room to leave with, and build everything else around getting them there: the problem that makes the claim necessary, the evidence or story that supports it, and what the audience should do differently once they believe it. A talk trying to cover five things the call for papers hinted at ends up covering none of them properly, because a room cannot hold five threads at once.

A structure that carries the argument

A working shape for a twenty to thirty minute slot: state the problem in under two minutes, build the argument through two or three sections, each landing its own small point, and close by returning to the opening problem with the answer now in place. Save exceptions and edge cases for the questions afterwards. If someone in the room needs the caveat, they will ask for it.

Timing that survives the room

A rehearsed run and the actual room are not the same thing, because the room has laughter, a slide that takes a second to load, and the odd moment where you lose your place. Build the talk in blocks you can drop if you are running long, and know which minute you should be at which slide before you walk on. A story that illustrates the second point but is not the second point itself is the first thing to cut.

A speaker still on slide four with five minutes left either rushes the ending or runs into the next speaker's slot, and the organiser then has to fix that in real time. Neither is where you want to be standing.

Slides that help rather than hinder

A slide should do one job the spoken words cannot: show a chart, a photograph, a diagram, a quotation worth reading. A slide full of bullet points repeating what you are saying gives the audience a choice between reading and listening, and most people read, which means they stop listening to you.

• One idea per slide, so nobody at the back has to decide what to look at first.
• Text large enough to read from the back of the room, which on most projectors means fewer words than feels natural while you are building the deck.
• A version of the talk that survives the projector failing: know the argument well enough to give it without slides at all.

Rehearsal that changes the talk

Running the talk through in your head is not rehearsal, because it skips the part where a sentence that read fine on the page falls apart out loud. Rehearsal means standing up and saying the whole talk at the pace and volume you would use on the day, more than once, and at least once in front of someone who can tell you where they got lost.

Time every rehearsal and note where it ran long. If the same section is slow every time, the section is too long.

Ready, or not yet

A talk is ready when you can say the closing line without looking at a slide, when you already know what you would cut if the clock ran short, and when you have said the whole thing out loud more than once. A talk that exists only as a slide deck and a plan in your head is not ready, however many times you have thought it through, because thinking a sentence and saying it out loud draw on different skills.

For a worked example of a talk broken into timed sections, see Worked Examples and Templates.

A talk that goes well in the room still needs following up properly, and this is the stage most first-time speakers skip because the adrenaline has gone and the next item on the agenda has already taken over. Getting selected once is not the same as being asked again, and the difference is usually decided in the hour after you leave the stage.

Handling questions without losing the room

Q&A is the part you cannot script, which is exactly why it needs a plan. Repeat or paraphrase the question before answering it, because half the room will not have heard it and you buy yourself a second to think. If you do not know the answer, say so and offer to follow up. Keep individual answers short: a long answer to one question eats time that belongs to everyone else waiting to ask theirs, and the organiser chairing the session will thank you for watching the clock.

A hostile or off-topic question is rare but not unheard of. The organiser or moderator should intervene if a question becomes a monologue or turns personal, and it is reasonable to ask them to before the session starts what they will do if that happens. This is also where a conference's code of conduct matters in practice: if a question or comment crosses into harassment, that is the organiser's problem to handle.

What to send afterwards, and when

If you promised slides during the talk, send them. Strip out anything confidential to your employer before you do, and decide whether you are sending a clean version built to be read without you narrating it, or the deck you actually presented with, which usually needs more context to make sense on its own. A short covering note with your slides and contact details does more for future opportunities than the slides alone, because it gives the organiser and anyone in the audience an easy way to reach you.

Many conferences record sessions. If yours did, ask the organiser when and where the recording will go live and whether you are allowed to share it yourself once it does. A recorded talk becomes the single most useful thing you can attach to a future proposal, because it lets the next reviewing committee watch you speak.

Getting feedback that is actually useful

Organiser feedback after a conference is often thin, a line or two in a thank-you email at most, so ask a specific question if you want a specific answer: how did the session compare on attendance and engagement to others in the same slot, and would they want a proposal from you again. Audience feedback, where a conference collects it, tends to be generous and unspecific unless something went wrong. Neither source will tell you as much as your own honest read of the room during the talk, which is worth writing down within a day or two while you still remember what landed and what did not.

Turning one talk into the next one

A single accepted talk is the strongest evidence you have for the next call for papers (CFP), so use it. Reference the recording or the slides in future proposals, mention the conference by name and the size of the audience if it is relevant, and ask the organiser directly whether they know of other events looking for a similar talk. Speakers at the same conference are also a route in: a conversation in the speaker room about where else you have both applied often surfaces events that never crossed your feed. None of this replaces writing a good proposal each time, but it does mean the second one is easier to write than the first.

For a template covering-note to accompany slides, and a sample proposal you can adapt for a second CFP, see the worked examples and templates page. If you are still deciding whether the speaking route is worth pursuing at all, the rest of the speaking at a conference guide covers what reviewers look for and how to prepare a talk before you get to this stage.