Many master's students do their thesis with a company: the company brings a question and data, the university brings the academic frame. It works well when both sides know their role. This guide covers the two supervisors, who assesses the thesis, what the agreement should settle, why the thesis is usually public, and how an academic thesis differs from the report a company would like to receive.
No one sees their own blind spots, with or without AI help. Esmond points them out in your draft from within your field, along with factual errors and inconsistencies.
A company thesis usually has a supervisor at the university and a contact person or co-supervisor at the company. The standard agreement of the Norwegian University of Science and Technology (NTNU) states the division plainly: “The responsible supervisor at NTNU has the overarching academic responsibility for the design and approval of the project description and the student’s learning.” The company side has its own duty: “The external organization must provide a contact person who has the necessary expertise to provide the student with adequate guidance in collaboration with the supervisor at NTNU.”
Karolinska Institutet (KI) in Sweden describes the same split in its own words. The academic supervisor or course leadership approves the topic and checks that the work can count as a degree project. The contact person at the company helps with access to information and gives feedback from the company's point of view. KI also advises agreeing early on how often you meet and how feedback reaches you.
In practice the two supervisors read your draft for different things. The company supervisor checks whether you understood the business, used the data correctly, and answered a question that matters to them. The university supervisor checks whether the question is a research question, whether the method fits it, and whether your conclusions follow from your evidence.
A hypothetical example of a commentYour methods chapter says the analysis uses twelve months of order data, but the results only cover the six months your agreement lets you show. Does the reader learn why the other six are missing?
A first comment on one part of your own draft, right away and without an account.
Get a free previewThe university assesses the thesis, under its own rules. The company can give input along the way, but the decision on whether the work meets the requirements of the degree is academic. Aalto University School of Business in Finland adds a detail that surprises some students: extra reports or confidential documents you hand to the company are not taken into account when the thesis is evaluated. Only the thesis counts.
The thesis therefore has to stand on its own. KI puts it as a condition: the thesis must contain enough information to allow assessment, and information that cannot be shared with the university cannot be used as a basis for supervision or examination. A finding you can only explain with data you are not allowed to show is, for the assessors, a claim without support.
| Person | Usual role in a company thesis | What to agree on early |
|---|---|---|
| You | Plan, carry out, and report the research; keep both sides informed | Timeline, meeting rhythm, who receives drafts |
| University supervisor | Approves the topic, guards academic quality, follows your learning | Whether the company's question can become a research question |
| Company supervisor or contact | Gives access to people and data, gives feedback from the business side | What data you may use and show, and how fast they reply |
| Assessors or committee | Decide on the thesis under the university's rules | Nothing directly; ask your supervisor how assessment works in your program |
NTNU makes a standard agreement mandatory when an external organization is involved. At KI and Aalto the university is not a party: any agreement is between you and the company. All of them want the terms settled before the work begins. At Aarhus University (AU) in Denmark the starting point comes from the national examination rules: “the student, the university and the third party must agree on the extent to which the third party's information may be used.” The NTNU agreement is signed by the student, the supervisor, the department, and the company.
Read the terms before you sign. Under the NTNU agreement, the starting point is that you own the results of your own student work. The company normally gets a non-exclusive right to use the results, and ownership passes to the company only if the agreement says so, or when the company claims a patentable invention under the terms of the agreement. Chalmers University of Technology in Sweden warns students to check whether a proposed agreement makes them give up their rights to their results, including the right to make them public, because the work has to be publicly accessible for approval. It also notes that “Chalmers does not negotiate the terms of agreements affecting students.” At AU, students who want advice on a company agreement have to find it outside the university.
At many Nordic universities a master's thesis is in principle a public document. Aalto's Systems Analysis Laboratory is direct about it: “According to Finnish legislation, a master's thesis is a public document and its contents cannot be confidential.” Aalto School of Business says the same for its programs: “All theses must be made publicly available, and an external party cannot demand that the thesis itself remains confidential.” Chalmers adds that company information in a report may be treated as confidential if the conditions of Swedish secrecy law are met, but “the material that forms the basis of an examination must always be public.”
KI draws the practical conclusion: “If a company wishes to keep large parts or the entire thesis confidential, the project is likely not suitable as a degree project.” Universities differ in the room they leave. NTNU allows delayed publication: “In special cases, the parties may agree that all or part of the work will be subject to delayed publication for a maximum of three years.” AU can sign a confidentiality agreement if it finds the terms acceptable, after which the assignment is labeled confidential and any oral examination is closed to the public. Check what your own university allows before you promise a company anything.
The usual solution lies in how you design the study. Choose a question that does not depend on trade secrets, report figures in relative or aggregated form, and leave customer names out. Agree on this with both supervisors before you collect data, because changing it after the analysis means redoing parts of the thesis.
Companies often want a short report with recommendations. You can give them one, but it is a separate document. KI says that a practice-oriented summary for the company “can be provided as a complement but does not replace the academic thesis.” Aalto asks the supervisor to approve the topic and to strike a balance between academic standards and what the company expects.
The difference shows in almost every chapter. A company report starts from the problem and ends with advice. A thesis also has to place the problem in the literature, justify the method, discuss what the data can and cannot show, and say what the findings add beyond this one company. Recommendations can still have a place, often as practical or managerial implications, but they need to follow from the analysis.
| Question | Company report | Master's thesis |
|---|---|---|
| Where does it start? | A business problem | A research question built from that problem and from the literature |
| What counts as evidence? | Whatever convinces management | Data and sources the assessors can follow |
| What about limitations? | Often left out | Discussed openly, including limits set by the agreement |
| Who is it for? | The company | The academic community, assessed by the university |
| How does it end? | Recommendations | Conclusions that answer the research question, then implications |
This example is made up. A master's student in supply chain management does her thesis with a mid-sized logistics company. The company wants to know why deliveries to one region are late. She interviews six planners, studies a year of delivery data, and ends with eight recommendations. The company supervisor is pleased.
Her university supervisor is less pleased. The literature review describes lean logistics in general but never returns in the discussion. The interview guide is not in the appendix, so a reader cannot see how the themes were found. Two recommendations rest on cost figures that the confidentiality agreement does not let her show. Nothing is wrong with the work itself; it is aimed at the wrong reader. A revision might connect the late deliveries to a concept from the literature, show the cost figures as indexed values the company approves, and add a section on what the case says beyond this one company.
Points drawn from the sources above. The last item is our suggestion.
Esmond evaluates every assignment before you submit. It never writes it.
Esmond is an AI tool that doesn't write for you. If your course doesn't allow any AI tools, don't use Esmond for that work. If you're unsure, ask your instructor or advisor.
Check your confidentiality agreement first. Chalmers reminds students that they are responsible for not passing a company's confidential information to third parties without the company's consent, and an online review tool is a third party. If the draft contains information the agreement covers, ask the company before you upload it, or upload a version without that material. Your report, which contains your full text, is erased automatically 90 days after your upload, and the AI model providers process data in the United States; the Privacy Notice lists each of them. What happens to your file is described on why a European review and in the Privacy Notice.
Esmond reads the draft you made and flags factual errors, inconsistencies, and blind spots, such as a methods chapter that promises data the results never show, or recommendations that go further than the analysis. It adds no text to the thesis. It doesn't know your company, your agreement, or the rules of your university, and it doesn't apply the rules of any country or program. The review is in English. See how Esmond works and the sample report.
Usually not in full. Aalto and KI say the thesis is a public document, and an external party cannot demand that it stay confidential. Some universities allow exceptions: NTNU allows delayed publication for up to three years in special cases, and AU can label a thesis confidential under a signed agreement. Check your own university's rules.
The university assesses it under its own rules. The company supervisor gives guidance and feedback from the business side. Aalto adds that extra reports you give the company are not taken into account in the evaluation.
It depends on the agreement. Under NTNU's standard agreement the basic principle is that the student owns the results, and the company gets a right to use them. Read the terms before you sign, and ask your supervisor if something is unclear.
Only if you agree to. Many students do. KI says a practice-oriented summary can be given as a complement, but it does not replace the academic thesis.
Read on November 7, 2026.
Get your paper evaluated before you submit