Conferences & Startups

Research Presentation

A 15-minute slot at the conference, then five minutes of discussion. Your paper holds two years of work, 40 pages, 87 references, and the room has already heard eleven talks today. eloqole gives your research presentation the structure that makes your finding land, even as talk number twelve.

Write my speech → start for free · exactly your speaking time

Last updated July 9, 2026

What a research presentation has to do

A research presentation is a talk of usually 10 to 20 minutes that prepares a research finding for an expert audience. The structure follows six stations: greeting with a hook, research question, key finding, method, data with interpretation, invitation to discuss. Everything is carried by a single core statement.

The difference from a business presentation is the burden of proof. Rhetoric advice from leadership seminars aims at effect; in the lecture hall, every number needs a source and every insight needs a method behind it. You are still allowed to sharpen. A talk is not a written publication: whoever just reads out what is in the proceedings anyway wastes their 15 minutes.

For researchers, the talk is also reputation work. Whoever presents clearly at an academic conference gets cited, invited, recommended. Graduate schools and career workshops treat presenting as its own craft for that reason, learnable like writing.

The structure: IMRaD translated for the talk

Papers worldwide follow the IMRaD scheme: Introduction, Methods, Results, Discussion. For the talk, this order only half works. In the paper, the finding may sit on page 28; in the talk it belongs in minute two: whoever knows the result can follow the derivation. Translated, the sequence looks like this:

1. Greeting and hook. One sentence about yourself, then a number or scene that shows the problem. The title on the slide may be dry; your first sentence may not.

2. Research question and relevance. What you wanted to know and why it moves the field forward. One sentence per point is enough.

3. The key finding. Your result, before the details: “We analyzed 1,200 patient records — and the effect everyone expects disappears as soon as you control for age.”

4. The method. The room needs three things: who was studied, how it was measured, why the design can answer the question. That fits into 90 seconds. Details move to backup slides for the discussion.

5. Data and interpretation in the main part. One number per slide, large, with one sentence on what it means in practice. Show relationships as a graphic; nobody reads a regression table with 30 cells on a screen. Name yourself what the study cannot deliver; two sentences of limitation take the edge off the most critical question.

6. Core statement and invitation. The close repeats the one sentence the room should keep and opens the discussion deliberately: “What interests me most: whether anyone has replicated the effect in clinical populations.”

The through-line is the core statement. Write it down before the first slide; any content that does not support it gets cut. A talk with two messages has none. The appearance succeeded if the room can still retell the core statement that evening.

Length and timing

The conference standard: 15 minutes of talk, 5 minutes of discussion. That is around 1,900 spoken words and at most 15 slides, better twelve. In a departmental colloquium, 30 to 45 minutes are common; for a poster pitch, two to three.

Hit the time window to the minute. Whoever runs over takes time from the next presenter, and every session chair remembers it. Cutting happens in preparation, never live, and you cut method and detailed data, never the finding and the close.

The most important preparation tip: a practice talk with a stopwatch in front of two colleagues, at least one from a neighboring discipline. You notice where you stall, and you get the questions in advance that are guaranteed to come up in the room. Rule of experience: the real talk runs ten percent longer than the rehearsal, because you slow down once it counts.

Versions: conference talk, colloquium, poster pitch

The conference talk. 15 minutes in a session with five other papers, a mixed expert audience, a hard time limit. Compression wins here: one finding, cleanly derived, told with grip. Different laws apply to the invited 45-minute talk: the keynote lives off a thesis and needs its own dramatic arc.

The colloquium. In a research or departmental colloquium, the audience knows you, and methods questions are welcome. More depth in the main part is allowed; the ground rule stays: finding early. If it is about your thesis, the exam format including the question round is covered on the thesis defense page.

The poster pitch. Two minutes next to your own poster, standing, against the noise of the coffee break. One finding, one graphic, one question to the person in front of you. More does not fit, and more is not the point.

What matters in the writing

Write spoken language. The paper sentence with three subclauses is precise in writing and convoluted out loud. Short main clauses, active verbs, technical terms only where the whole room knows them. The test: can the colleague from the neighboring discipline follow every sentence? Speaking clearly costs no precision; the exact values are on the slide.

Bullet points on the slide, the script in your head. A free delivery looks composed. It grows out of a fully written text that you rehearsed aloud until bullet points suffice as memory aids. Whoever speaks straight from bullet points meanders; whoever reads aloud loses the room.

Ration the visual aids. One chart beats any table. Short videos wake the audience up but need tested audio. Tip: play them on the venue computer beforehand, never only on your own. Hand out materials after the talk; before it, the paper competes with you for attention.

Sources sparingly, but cleanly. Only scholarly publications count as evidence; the newspaper article works at most as a hook in the first sentence. The slide gets the short form with author and year, the full reference list goes in the handout.

Facts need meaning. “Odds ratio 2.3” tells the room nothing; “the risk doubles” sticks. Translate every central number into one crisp everyday sentence.

What a complete talk opening and a slide sequence with one core sentence per slide sound like is shown in the research presentation examples, written out and annotated.

The most common mistakes

Four mistakes ruin almost every research talk, and all four are avoidable:

Overstuffed slides. 200 words per slide force the room into a choice: read or listen. Nobody can do both at once. One statement per slide; the rest is script.

Reading aloud. Whoever clings to the manuscript loses eye contact, gesture, and pace; listeners check out after two minutes. Speaking freely is a result of rehearsing, see the practice talk.

Jargon in front of a mixed audience. At interdisciplinary conferences, statisticians sit next to practitioners. Researchers chronically underestimate how specialized their vocabulary is: every unexplained abbreviation costs part of the room, three of them cost the discussion.

The method eats the talk. Eight minutes of sample description, then the findings race past in three. Flip the weighting: 90 seconds of method, the rest for findings and interpretation.

That leaves stage fright. A shaky voice at slide one is normal and trainable. The guide to overcoming stage fright shows the routines that work before you go on.

How your talk comes together with eloqole

You give eloqole your research question, the key finding, the essentials of the method, and your time window. Out comes a talk structure with the finding up front, then the fully written script for your slides, in the tone of your discipline. You adjust the terminology and rehearse in the teleprompter until the 15 minutes stand exactly.

1

Tell

Keywords, names, moments — eloqole asks the right follow-up questions, rough notes are fine.

2

Shape

Pick tone and speaking time. Rearrange the outline until it fits.

3

Deliver

Read the finished speech, refine it and rehearse with the teleprompter until it sticks.

Frequently asked questions

+How is a research presentation structured?

Six stations: a greeting with a hook, the research question with its relevance, the key finding, the method in 90 seconds, data with interpretation, a close with the core statement and an invitation to discuss. That is the IMRaD logic of the paper, rearranged for the talk: the finding moves to the front.

+What are the 10 golden rules of presentations?

The lists circulate in many versions. What holds up in the lecture hall: one core statement, finding early, one number per slide, slides as images instead of a script, speak freely, keep eye contact, respect the time, ration the jargon, rehearse out loud, end with a question to the room.

+What is the 10-20-30 rule for presentations?

Ten slides, 20 minutes, no font below 30 points: a rule of thumb from investor Guy Kawasaki, invented for startup pitches. For a research presentation it works as the lower bound of readability. The slide count here follows the slot: one slide per minute, ideally fewer.

+How much content fits into 15 minutes?

One finding, cleanly derived. As a rule of thumb, one slide per minute, ideally fewer. Whoever tries to fit the whole paper rushes, and the room checks out at slide six. eloqole writes the script to your exact time window.

+Do I have to explain the full methodology?

No, only what is needed to trust the finding: sample, design, the central instrument. Details like item wording belong in the appendix and in the discussion, in case someone asks.

+Should I raise the weaknesses of my study myself?

Yes, briefly and before the discussion. Whoever names the small sample or the missing longitudinal data first takes the edge off the most critical question. Two sentences of limitation, one sentence on what holds regardless.

+How do I prepare for the discussion?

Write down the three questions that would annoy you most and draft full answers. Keep a backup slide with detailed data ready. A calm “We did not examine that, but it would be the next step” is a complete answer.

+Does eloqole also work for conferences in other languages?

Yes. You can have the talk written directly in English or enter bullet points in one language and get the script in another, including teleprompter practice.

Related occasions

Your first draft is waiting

Answer a few questions and read your first draft within minutes. Edit, refine and rehearse until it sounds like you.

try it for free →