How to Make a Thesis Proposal Presentation: 10-Slide Structure

A thesis proposal presentation condenses a proposal document of several thousand words into about ten slides that you deliver in a short talk, usually followed by questions from your committee. It has to answer four questions: Is the problem worth studying? What does existing research already cover? How exactly will you do the work? Can you finish it in the time you have? Order your slides to follow those questions (background, literature review, aims, methodology, contribution, timeline, expected outcomes, references) and spend most of your speaking time on the literature review and the methodology. Below: a 10-slide structure with timings, talking points for the key slides, likely committee questions, and an AI workflow that starts from your Word document.

What the presentation has to cover

Programs call this milestone different things: proposal defense, prospectus, qualifying talk, confirmation or upgrade review. The content, though, tends to fall into the same eight parts:

• Background and significance. What real problem exists, and why solving it matters in theory or practice.
• Literature review. The main strands of prior work, what each has achieved and what is still missing.
• Research questions and objectives. One overall aim broken into two to four concrete objectives.
• Methodology and research design. How you will tackle each objective, and how the steps connect.
• Contribution. What is new about your angle or method compared with existing work.
• Timeline. What happens in each phase between now and submission.
• Expected outcomes. Thesis chapters, papers, datasets, software or reports, whichever genuinely apply.
• References. A short list of the core sources, in one consistent format.

You have no results yet. What you are proving is that the plan is clear, the method is feasible and the schedule is realistic. That is why the literature review and the methodology carry the talk, and why the background should stay short.

A 10-slide structure for a 10-minute talk

Use this as a skeleton and adjust it to whatever your department asks for:

A few adjustments are common:

• If your literature is broad, split slide 3 so each strand gets its own slide.
• In the humanities and social sciences, slide 6 is often a conceptual framework rather than a technical pipeline. It should still be a diagram.
• Some programs expect a slide on ethics approval or research resources. Add it if your guidelines mention it.

How many slides, and how long?

Practice varies a great deal between universities, departments and countries, so the instructions you receive always take priority. As a rough guide, proposal talks often run somewhere between 5 and 20 minutes before questions, and one slide per minute is a sensible starting estimate:

Whatever the length, the budget stays the same. Background takes no more than about a fifth of your time, and the literature review plus the methodology take roughly half. Spending four minutes on context and then rushing the method is a common mistake.

Writing and presenting the key slides

Background: two sentences, not a history lesson

Skip the history of your field. State the real-world problem in one sentence and why it has not been solved in the next. For example: "Identifying crop disease in the field still relies on expert inspection, which is slow. Current deep learning models are too large to run on the phones farmers actually carry." After those two sentences, the committee knows what you are trying to fix.

Literature review: group by approach, not by year

The usual mistake is a chronological list of everything you have read. A clearer approach:

1. Identify two to four strands of related work, for example handcrafted image features, general-purpose deep models and lightweight models.
2. Give each strand a one-sentence verdict on what it achieved, with one or two representative citations.
3. Use a separate slide for the gap, phrased along the lines of "Work on X is well developed, but Y under real-world conditions remains open."
4. When presenting, let the gap lead directly into your research question.

On this slide the committee is really checking whether you have read the work you cite. Be ready to explain the core method of any paper that appears on screen.

Methodology: let the diagram do the explaining

The objectives slide lists your two to four tasks, one sentence each. The methodology slide connects them in a single diagram: the research question on the left, phases such as data collection, method design, evaluation and analysis across the middle, the specific technique labeled in each box, and the output of each phase on the right. Arrows show order and dependency. When you present it, walk the arrows and say one sentence per box: what you will do and what it produces.

Contribution and feasibility: modest beats grand

Keep contributions to two or three, each stating how it differs from existing work. For feasibility, list concrete conditions: whether you can access the data, whether the equipment is available, what groundwork your lab has already done. Avoid promising a specific accuracy or effect size at this stage. An unsupported number is an invitation to a hard question.

Timeline: plan backward from submission

Start from your submission or defense date and work backward by month or semester. Leave a solid block for writing and revision. A Gantt chart or milestone table is far easier to read than a paragraph.

Questions committees often ask

Proposal questions tend to focus on feasibility. Prepare short answers to these:

• How does this differ from the work you reviewed?
• Is the scope realistic for the time you have left in your program?
• Where will the data come from, and can you get it on schedule?
• Why this method, and what alternatives did you consider?
• If a key step does not work, what is your fallback?
• What exactly did a particular paper on your review slide do?

Three delivery habits help. Lead each slide with its conclusion instead of reading the text aloud. Write down every suggestion the committee makes rather than arguing on the spot. And if you cannot answer a question, say that it needs further reading. That lands better than an improvised answer.

Design rules: one idea per slide, method as a diagram

A proposal deck does not need decoration. Clarity is the design:

• One idea per slide. Each slide answers one question, and its title states the conclusion, such as "Lightweight models lose accuracy against cluttered field backgrounds," rather than "Related work."
• Method as a diagram. Use a flowchart or block diagram with short phrases in the boxes and arrows running in one direction.
• Timeline as a chart or table. Time on one axis, tasks on the other.
• Sentences become fragments. A paragraph in your proposal usually becomes two or three short lines on screen. The full argument stays in the document.
• Template first. Use the department template if there is one. Otherwise, one accent color and one font family across the deck.
• Readable references. A smaller font is fine, but it has to be legible from the back of the room.

From Word document to editable slides with AI

Once the proposal is written, splitting it into slides and laying them out is mechanical work that AI can handle. In our AI slide studio the workflow looks like this:

1. Clean up the document. Remove the cover page, form instructions and formatting notes. Give each part a clear heading such as Background, Literature Review, Methodology and Timeline.
2. Upload or paste. Upload a .docx, .txt or .md file, or paste the full text (up to 100,000 characters). PDF upload is not supported, so copy the text out of the PDF and paste it.
3. Add instructions. For example: "10 slides, clean academic style, methodology as a flowchart, timeline as a table." You can also choose the output language, which helps if you wrote the proposal in one language and are presenting in another.
4. Review the outline. The AI first drafts a slide-by-slide outline of titles and key points, and creating it costs 0 credits. Compare it with the 10-slide structure above, then add, delete or reorder slides. What you confirm is exactly what gets generated. Our guide to outline-first presentations explains why this step matters.
5. Pick a style and generate. Choose a clean academic look from more than 100 styles. The cost is shown before you generate: 2 credits per slide with AI images included, so a 10-slide deck is 20 credits, and it usually takes one to three minutes. A lab can save its logo, footer and preferred style in a brand kit and reuse them.
6. Export and refine. Download an editable PPTX and open it in PowerPoint or WPS to apply your department template, adjust the method diagram and check each reference. For why editability matters, see editable PPTX vs image slides.

The result is a high-quality, editable first draft whose content still comes entirely from your own proposal. The AI handles structure and layout; the research question, sources and study design stay with you and your supervisor.

To see what an academic deck looks like in practice, open the master's thesis defense example. It includes the original input, so you can see how a set of research notes became 12 slides. For the full path from proposal to final defense, visit our AI thesis defense slides page. When your proposal is ready, upload it in the studio and start with the outline.

FAQ

How is a proposal presentation different from a thesis defense?

A proposal presentation is about the plan: why the question matters, what has already been done, how you will do the work and whether the timeline is realistic. A defense is about what you found. At the proposal stage you have no results yet, so do not present target figures as if they were data.

Should I use my university template or pick my own design?

If your department provides a template, use it and fill in the title slide as instructed. If not, a clean academic look works well: a white or light background, one accent color and one font family. The committee cares much more about a clear research question and a feasible method than about the template.

Can AI write the literature review slides for me?

Use AI to split and lay out the content you have already written, not to supply sources. Every reference on your slides should come from your own proposal, and you should check authors, years and venues one by one. Check your university's rules on AI tools before you start.

My proposal is a PDF. Can I upload it directly?

PDF upload is not supported. Upload a .docx, .txt or .md file instead, or copy the text out of the PDF and paste it. Pasted text can run up to 100,000 characters, which is plenty for a typical proposal.