How to Write a Master's Thesis in Germany: Outline, V0, Feedback, and How It Is Graded

October 3, 2026 · Career Advice · 11 min read

By Dr.-Ing. Faisal Qayyum. Stuck with your own thesis? See how I can help.

Hari had four chapters planned and data for three of them. He was still on chapter two.

He was a Master’s student in metal forming at TU Bergakademie Freiberg, writing his thesis on electromagnetic forming, and he had decided that chapter two had to be perfect before chapter three could start. By the time it felt good enough, the clock had eaten the weeks he needed for everything else. When we recorded an Offload Podcast episode about thesis writing, that was the first problem he brought to the table, and I recognised it instantly. I spent nine years at the Institute of Metal Forming in Freiberg supervising and grading theses like his. Almost every student I worked with hit some version of it.

This post is the written version of that conversation: the method I give my own students, and what the supervisor is actually looking at when the grade is decided.

The full episode (62 minutes). The video loads from YouTube only when you press play.

A Good Thesis Is a Done Thesis

The only characteristic of a good thesis that matters on submission day is that it exists, finished, on the table. Not that every figure is beautiful or the English flawless. Students rarely fail to finish because they lack ability. They stall because of a handful of beliefs that feel reasonable and quietly stop the writing.

I learned this the hard way. When the pandemic hit, I had a pile of results and suddenly a lot of free time, and I still could not write. Time was never the problem. These beliefs were:

Hari’s chapter-two problem was the last one in disguise. Perfectionism looks like high standards. In practice it is a way of never reaching chapter three.

Build the Outline Before You Write a Sentence

An outline turns writing into filling in blanks. It shows you every gap you have to fill, so you always know what to work on next. Make it as detailed as you can, in your own words, then check it with your supervisor, your colleagues, and two or three earlier theses from your group.

This is less work than it sounds, because thesis structure in engineering is fairly standard: introduction, state of the art, problem statement, objectives, methodology, results, discussion, conclusions. Under each heading, write a few bullets on what will go there. Hari described what this did for him better than I can: before the outline, the thesis felt like a desert with no road; after it, every section had a signboard pointing the way.

The outline will change as the work changes. That is fine. You only need something to start from.

Write in the Order the Work Happens

Nobody has to write a thesis in the order it is read. The order that causes the least friction follows the work itself:

1. Problem statement. It is on your table before you do anything else, so write a rough version on day one.
2. Methodology. You know what you clicked, cut, tested and measured. Write it while you do it.
3. Results, a few at a time, as they arrive.
4. Introduction and state of the art, while you compare your results with the papers you are reading anyway.
5. Discussion and conclusions, as the patterns become clear in conversations with your supervisor.
6. Objectives, abstract, keywords and title last.

Write with the end in mind. Everything you produce, every bullet, sketch and figure, will end up in the finished thesis in some form. The habit I push hardest with my students: keep one PowerPoint file from the first day, and add a slide every working day with what you did, what you read and what you found. After a few months it becomes a diary of your whole thesis, and copying from it into your outline is far easier than reconstructing six months from memory.

Stop Producing Data and Talk Your Thesis Into a V0

At some point your outline will hold bullets, rough figures, references and notes from your supervisor. That is the moment to stop producing new graphs and start the first full draft. I call it V0, version zero: every thought on the page, readable by you, not yet presentable to anyone else.

The fastest way I know to write it is to talk. I put my phone on record, open the page with the figure, and explain it out loud as if to a friend: what the colours mean, where the trend goes, how it compares with the literature. Then I play the recording back and type what I said, pausing whenever an idea for an extra arrow or a better comparison comes up. Thinking and writing at the same time is hard; talking first and typing second splits them. Speak in whatever language your thoughts flow best in. The English can be fixed later; the story cannot be recovered if you never got it out of your head.

A dictation tool removes the typing too: speak in English, and a 15-minute recording becomes text in a couple of clicks. I described the same V0 approach for research papers in How to Use AI to Write a Paper (Without the Fake Citations).

From V0 to V1: Edit, Then Climb the Feedback Ladder

Turning V0 into V1 takes real time. You go through every line, correct it, add the figure you promised yourself, find the reference you flagged. But the story is already on the page, so you are editing, not inventing, and editing is a far easier problem.

Then get feedback, and climb the ladder slowly. Ask a colleague, a friend, even a parent with some technical background, to read five or ten pages and tell you where they got lost. Fix that. If someone outside your topic can follow two representative pages, the rest of the thesis written the same way is in good shape. Only then go to your supervisor. Each round moves you up a version, and there is no upper limit except the time you have.

How Your Master’s Thesis Is Actually Graded

This is the part most students never see. At the Institute of Metal Forming in Freiberg, every thesis I supervised was graded on a written rubric, and three people filled one in: the supervisor and two examiners. Rules differ between universities and departments, so check your own examination regulations. Here is what our rubric looked at:

Two details matter more than students expect. First, the two examiners only read your thesis, while your supervisor watched you work, so the supervisor’s assessment carried more weight. When the scores differed a lot, they were averaged, because that gap says something: either you did excellent work that never made it onto the page, or you wrote well about work your supervisor did not see you do. The best result is when both agree. Second, the written thesis was only half of the final grade at our institute. The presentation, graded on its own rubric, was the other half.

Under every rubric, the assessors kept returning to the same questions. Did you define the problem clearly? Did you show the research gap in the earlier work? Were your objectives clear? Could they follow your methodology as a story? Did you explain your results and compare them honestly with the literature, including what did not work? Did you state what is new?

The lesson is uncomfortable but useful. The polish students lose sleep over is the smallest part of the grade. Understanding your own work, and showing that understanding on the page, is most of it. If your supervisor knows you did good work but the thesis does not show it, you lose marks you had already earned.

Tools That Save Weeks, and the One Rule for AI

Investing a little time in tools pays back many times over, and the time you spend learning them counts as writing time too. The ones I recommend most often:

• A citation manager from the first day, so the reference list never becomes its own project.
• Literature tools: Elicit to find relevant papers, Litmaps to see clusters of work you might be missing, typeset.io to help you read dense papers and equations.
• A language checker such as Grammarly; the free version covers most needs.
• Dictation, built into Word and many other tools, for getting V0 out of your head.
• Generative AI to turn your own rough V0 text into clearer paragraphs.

The one rule for AI: it works on your material, never instead of it. Your thoughts, your results and your sources go in; clearer sentences come out. Check your university’s rules on AI use before you start, declare it where they require, and never let a tool supply facts or references you have not checked yourself.

The other tool is people. If a friend is good at Excel or MATLAB, ask for help. If someone in your group made beautiful figures, ask for their template. Read a colleague’s chapter when they ask, and they will read yours. Learning to outsource what others do better is a skill you will need long after the thesis.

The Habits That Finish Theses

Give time to writing instead of finding time for it. Block the same slot every day, even if it is only 30 minutes. I often suggest waking an hour earlier and writing before anything else. Some days you will stare at the page; that is fine, come back tomorrow. Once V0 exists there is always a bullet to expand or a figure to fix, and most days the 30 minutes turn into more.

Keep reading. When students start writing they stop reading, and the ideas dry up. You cannot improve a figure if you have not seen a better one recently.

Find your own method. Not every technique here suits everyone. Some people think best out loud, some on a walk, some with a notebook and a pen. Hari told me he could listen for hours but found long reading hard, so recording suited him. Try a few and keep what works.

Expect stress, and manage it. A thesis is meant to be demanding; a fixed deadline and a presentation are part of what is being tested. Before a thesis started, we asked students about their circumstances, including their finances, which some found oddly personal. We asked because we had learned that money worries and poor health show up in the thesis later. If you are struggling with either, tell your supervisor early, while something can still be done.

Where Are You Stuck? What to Do This Week

Final Thought

Your Master’s thesis is not the last thing you will write. It is the first formal document with real weight attached, and almost everything after it, papers, proposals, reports at work, will carry more. Hari pointed out that the first question in many job interviews is “what did you do in your thesis?” The process you learn now, outline, V0, feedback, versions, is the one you will use for the rest of your career.

So start rough, start early, and keep moving up a version at a time. There is no version minus one. There is only V0, and everything you build on it.

If You Are Planning What Comes After the Thesis

If the thesis is going well and the next question is a PhD or a job, PhD or Industry? What Engineering Students in Germany Are Not Asking Themselves is the conversation to have first. And if a PhD is the plan, How to Build a Postdoc Network in Germany Before You Need One explains why the people in your field should know your name before you apply.