University of Oulu Business Analytics Master’s Thesis Guide: 721020S, 30 ECTS, OBS Seminar, 721010S Maturity Test and Laturi

Quick answer: what is the University of Oulu Business Analytics Business thesis route?

The current Business Analytics Business path at the University of Oulu leads to a Master of Science (Economics and Business Administration) at Oulu Business School. The degree is 120 ECTS over two years, and the current 2026-2027 programme object is 51732. The exact thesis is 721020S Master’s Thesis, Business Analytics, 30 ECTS, classified as Advanced Studies and graded 1-5/FAIL. The 30 ECTS course itself includes the thesis seminar, planning, writing, presentations and reports. The degree also requires 721010S Maturity Test, Master’s Degree, Business, 0 ECTS, assessed pass/fail. Final thesis handling uses Laturi, while degree-certificate application is completed through Peppi.

A major boundary is important from the beginning. Business Analytics has three degree-specific paths. The Business route is not interchangeable with the Computer Science and Engineering or Software Engineering and Information Systems routes. This guide therefore uses only evidence that applies to the MSc Economics and Business Administration path or to all Oulu Business School Master’s students. The current public sources used here do not establish an exact Business-path examiner count, so no examiner count is invented.

1. Confirm that your degree path is Business, not one of the two technical paths

The Business Analytics programme combines business and technology, but students enter one of three degree-specific study paths. The Business route awards the MSc in Economics and Business Administration and develops Business Analyst competence. The Computer Science and Engineering route awards an MSc in Technology, while the Software Engineering and Information Systems route awards an MSc. Shared courses do not make their thesis administration identical. Before following any thesis instruction, verify that your Peppi programme is 51732 Business Analytics (MSc, Economics and Business Administration).

2. Read the current 120 ECTS structure before planning the thesis year

Programme 51732 places 92-94 ECTS in Advanced Studies in Business Analytics and also includes optional studies. The advanced-studies structure contains quantitative methods, data and decision-making, market and customer analysis, enterprise process planning, capstone work, responsible business, the 30 ECTS thesis and the 0 ECTS maturity test. The thesis is therefore a large part of the degree, but it should not be treated as the entire second-year workload. Use the current personal study plan to see which courses remain alongside thesis work.

3. The exact thesis is 721020S and it is 30 ECTS

The current programme structure lists 721020S Master’s Thesis, Business Analytics at 30 ECTS. The current implementation is 721020S-3006, running through the 2026-2027 academic cycle, and the course is taught in English. It belongs to Advanced Studies and uses the 1-5/FAIL assessment scale. These details should control the current guide. Do not replace 721020S with a similarly named thesis course from a technical Business Analytics path or another Oulu programme.

4. The seminar is embedded inside the 30 ECTS thesis course

The 721020S course description explicitly includes the Master’s thesis seminar, thesis planning, thesis writing, presentations and reports. Programme 51732 does not add a separate credited thesis-seminar object on top of the 30 ECTS thesis. For this path, it is therefore safer to describe the seminar as part of the thesis process rather than to invent extra seminar credits. If a future curriculum later separates the seminar, the current Peppi structure should be rechecked before using the new arrangement.

5. Oulu Business School uses group supervision as the normal model

The current Oulu Business School process states that the primary Master’s thesis supervision model is group supervision including seminar work. Group work normally starts in autumn, and a supervisor is appointed to each student. This arrangement matters because the thesis is not only a private sequence of meetings between student and supervisor. Progress is also developed through presentations, comments from the group and opponent work. Students should therefore plan the thesis timeline around the scheduled supervision stages, not only around a personal final-submission date.

6. The supervision process has four stages

Oulu Business School structures the process around topic selection, research plan, intermediate report and manuscript. Topic selection allows early discussion of possible ideas. The three later stages are developed through group and individual supervision. Treat these stages as decision gates. The research plan should establish the question and method; the intermediate report should show that theory and evidence are developing coherently; and the manuscript stage should test whether the complete argument is ready for final corrections and Laturi submission.

7. Your thesis topic needs supervisor approval before you begin

The current OBS instructions state that the thesis topic must be approved by the appointed supervisor before thesis work starts. This is more than an administrative step. A Business Analytics topic can easily become too broad, too technical, too descriptive or too dependent on unavailable company data. Before committing to a topic, make the problem, research objective, expected data source and intended analytical contribution clear enough for the supervisor to judge whether the project is realistic within a Master’s thesis.

8. Prepare the research plan as the first serious research design document

OBS describes the research plan as approximately five pages. Use those pages to define the business problem, academic question, theoretical perspective, data, unit of analysis, method, expected contribution and practical limitations. If the study depends on a company dataset, API, customer records or internal dashboard, state that dependency explicitly. A strong plan does not pretend that uncertain access is already secured. It shows how the research can proceed and what fallback scope is possible if the preferred data are delayed or unavailable.

9. The intermediate report should prove that the thesis is becoming a study, not just a project

The current process describes the intermediate report as approximately 30-40 pages, broadly corresponding to the theoretical part. At this stage, the literature review should already explain the concepts and prior evidence needed to interpret the analysis. The research question should be stable enough that the method can be justified against it. If the thesis is still only a description of a dashboard, company process or dataset, use the intermediate stage to identify the missing academic comparison, mechanism, relationship or evaluative criterion.

10. The manuscript stage should contain the complete argument once

OBS describes the manuscript as approximately 60-80 pages, with all chapters written once. The objective is not perfect language on the first full draft. The important requirement is that the whole reasoning chain exists: problem, literature, research question, method, evidence, results, discussion, limitations and conclusions. A complete manuscript lets the supervisor see contradictions that cannot be detected when chapters are reviewed in isolation. It also creates enough time to correct methodological problems before the final Laturi version is frozen.

11. Presentations and opponent work are part of the thesis process

Students are expected to present their own work and act as an opponent to another student’s thesis at least once. The opponent task includes an A4 report. Use this activity as methodological training rather than a formality. When reviewing another thesis, ask whether the research question matches the data, whether the method answers the question, whether conclusions exceed the evidence, and whether the theoretical framework is actually used in the analysis. Apply the same questions to your own thesis afterward.

12. Oral presentation is a separate learning outcome

OBS requires an oral presentation by the intermediate-report or manuscript stage so that oral communication can be evaluated. The current instructions also state that written and oral communication evaluations are separate from the thesis grade. Even so, presentation quality matters to the supervision process. A concise presentation forces the researcher to state the problem, method and result clearly. If the central contribution cannot be explained in a few minutes, the written argument may also need sharper focus.

13. An independent route exists, but it changes the supervision entitlement

OBS allows work outside the standard process only after separate discussion with the Head of Master’s thesis research. Under that route, the student gives up the normal process supervision and works more independently, although the final manuscript still needs presentation for oral-skills evaluation. This is not simply a shortcut around seminars. A student considering it should understand that less structured supervision creates more responsibility for methodological planning, deadlines, feedback and compliance with the final thesis requirements.

14. 721020S expects scientific research, not only business problem solving

The thesis learning outcomes require students to choose and apply appropriate research methods, produce new knowledge for Business Analytics problems, conduct scientific research and make reasonable recommendations. Practical relevance is therefore valuable, but it does not replace research design. A company may ask, “Which customers are most likely to churn?” The thesis must still define the population, outcome, data, comparison, model or analytical method, validation strategy, limitations and relationship to prior research.

15. Connect theory, empirical evidence and recommendation

The thesis course expects understanding of Business Analytics theories, concepts, frameworks and empirical findings. This means a recommendation should not appear only because a model or dashboard generated a number. The literature should explain why the variables, relationships or performance criteria are meaningful. The empirical analysis should then test, estimate or evaluate something that connects to that theoretical argument. The recommendation should follow from both the evidence and the boundaries of the study.

16. Choose the unit of analysis before running statistics

Business datasets often contain many rows without containing many independent cases. One customer may make hundreds of transactions, one firm may contribute monthly observations for years, and one campaign may generate thousands of clicks. Decide what the conclusion is intended to generalise to: customers, firms, transactions, markets, periods, campaigns, products or another unit. Sample size, statistical dependence and validation design should then reflect that unit rather than the number of rows in a spreadsheet.

17. Quantitative analysis should address validity, reliability and generalisability

The current 721026S Statistical Methods for Business Analytics course explicitly trains students to evaluate the usability, validity, reliability and generalisability of data and reports. These are useful thesis questions. Are variables valid measures of the concepts in the research question? Are results stable enough to be trusted? Does the sample support the population claim? Are missing data or selection processes likely to distort the conclusion? Method sections should explain these issues rather than only naming the software used.

18. Software is a tool, not the method itself

The programme’s statistics environment includes tools such as Excel, SPSS, PSPP and Mplus, while other courses introduce BI and enterprise systems. Writing “the data were analysed in SPSS” does not explain the analysis. State the model or statistical procedure, variables, assumptions, preprocessing, decision rules and diagnostic checks. The software version can be recorded for reproducibility, but the academic method must remain understandable even to a reader who uses a different package.

19. Enterprise data require provenance before interpretation

721025S Enterprise Process Planning focuses on combining, integrating, modifying and rearranging different data sources into meaningful information for analytical decision-making. A thesis using ERP, CRM, sales or process data should document where each field came from, how tables were joined, which records were excluded and how derived variables were created. If the analysis cannot trace a result back to the source data and transformation rules, the business recommendation becomes difficult to audit.

20. Cybersecurity and access controls can affect the research design

Enterprise Process Planning also includes business perspectives on cybersecurity. In a thesis, this does not mean every project requires a cybersecurity chapter. It means that access to organisational systems and data can shape what is ethically and practically possible. Do not move internal datasets into unapproved personal tools merely because analysis is easier there. Agree access, storage and permitted processing with the data owner and University guidance before building the analysis workflow.

21. Business Intelligence projects need more than attractive dashboards

813320A Business Intelligence: Applications and Projects covers BI systems, dashboards, visualisation, data warehousing, big data and organisational decision-making. If a thesis creates a dashboard, define the decision problem it is supposed to support. Explain data sources, refresh logic, filters, measures and validation. A visually impressive dashboard can still mislead if denominators change silently, categories are inconsistent or a calculated KPI has no defensible business meaning.

22. Make visualisations reproducible

For every important chart or dashboard figure, keep enough information to reproduce the result: source dataset version, filters, time range, grouping, calculated fields and code or configuration. If a figure is exported manually from Power BI, Tableau, Excel or another platform, record the state that generated it. This prevents a common final-stage problem where the thesis contains a chart that no longer matches the current dataset or the analysis described in Methods.

23. Predictive analytics needs a genuinely separate evaluation logic

If the thesis develops a prediction or classification model, decide early what future case the model is intended to predict. Training performance is not evidence of generalisation. Preprocessing, feature selection, hyperparameter tuning and threshold selection should be made using development data. A final hold-out set should remain outside those decisions if it is intended to estimate unseen performance. For time-dependent business data, random splitting may also leak future information into the past, so chronology may matter.

24. Cross-validation should match the business data structure

The programme includes data-mining competence, but no single cross-validation scheme is correct for every thesis. Customer-level prediction may require grouping all records from one customer together. Firm-level studies may require organisation-level separation. Time-series forecasting usually requires time-respecting evaluation. The thesis should explain why the chosen resampling design represents the deployment or inference problem. Repeating a standard k-fold routine without considering dependency can produce an overly optimistic result.

25. Report uncertainty and denominators with headline metrics

A model accuracy, revenue uplift, conversion rate or average effect is easier to interpret when the denominator and uncertainty are visible. Report how many customers, firms, periods or observations contributed to each analysis, how many were excluded and why, and how class or outcome distributions differ across partitions. Where appropriate, include confidence intervals, variability across folds or sensitivity analyses. A large transaction table can still represent a small number of independent organisations or customers.

26. Descriptive analytics and causal claims are different

A dashboard can show that two measures move together, and a regression can estimate an association, but neither automatically proves that changing one variable will cause the other to change. If the thesis uses causal language such as “impact,” “effect” or “drives,” the design must justify that interpretation. Otherwise, describe the result as association, prediction, comparison or pattern. Clear claim boundaries are especially important in business research because recommendations can easily sound more causal than the evidence supports.

27. Company collaboration does not remove the need for public thesis planning

Business Analytics is designed around real organisational problems, and Oulu Business School students may work with company data. Laturi guidance states that the actual Master’s thesis must not contain secret trade or professional material. Confidential background material must be handled separately and agreed in advance with the supervisor and, for commissioned work, the commissioning party. Resolve publication boundaries before analysis, not after confidential figures have already been placed in the manuscript.

28. Personal data require planning before processing begins

The University of Oulu’s current data-protection policy is explicit for theses. If the project processes personal data, the planned lifecycle of processing must be recorded in the research plan before processing begins. The plan must address data protection and information security, include a risk assessment and consider whether a DPIA is required. Research subjects must receive an appropriate privacy notice. Data should be minimised and anonymised or pseudonymised where possible.

29. Pseudonymised customer data are still personal data

Replacing names with customer IDs does not automatically make a dataset anonymous. If a key exists or a person can reasonably be re-identified from combinations of attributes, the data remain personal data. Business datasets can contain indirect identifiers such as exact timestamps, location, rare purchases, account history or job roles. Keep re-identification keys separate from routine analysis files and limit access according to the approved arrangement.

30. Ethics review depends on the actual study design

Not every Business Analytics thesis requires an ethics-committee statement. However, studies involving human participants, sensitive personal data, interventions, deception or other conditions covered by ethical-review guidance may require preliminary assessment. The University Human Sciences Ethics Committee handles applicable non-medical human-sciences requests. Classify the project early with the supervisor. If review is required, obtain it before the research activity for which the review is needed, rather than trying to regularise the project afterward.

31. Build a data-management plan that supports reproducibility

University responsible-research policy expects data-management planning to address collection, storage, sharing, preservation and reproducibility. For a Master’s thesis, a practical version can include a data inventory, variable dictionary, source and access conditions, transformation log, code location, output folders and retention/deletion plan. Agree ownership and access rights early, especially when company or externally licensed data are involved. The aim is to preserve enough provenance to audit the result without exposing data that should remain closed.

32. Research integrity applies to code, data and AI as well as prose

University ethical principles require students to respect the work of others and avoid presenting another person’s work or text as their own. In Business Analytics, the same discipline should extend to code, datasets, generated analyses and AI-assisted material. Keep track of external code, packages, prompts or generated content when they materially contribute to the work, and follow the current University and OBS instructions on AI reporting. Turnitin similarity checking cannot detect every methodological or authorship problem.

33. The maturity test is 721010S, 0 ECTS and pass/fail

Programme 51732 separately lists 721010S Maturity Test, Master’s Degree, Business, carrying 0 ECTS. Oulu Business School uses a standard Master’s maturity question asking the student to present thesis outcomes in relation to prior scientific research and discuss how the outcomes can be used in business life or economic decision-making. The test therefore checks more than memory. It asks the student to connect the completed thesis to its scientific and practical meaning.

34. OBS uses a personal electronic Exam for the maturity test

The current OBS instructions state that the maturity test is completed as an electronic exam in the camera-supervised PR106 room. It can be taken after the manuscript presentation has been accepted by the supervisor. The supervisor creates and publishes the personal exam in the Exam system, and the student reserves the time through the notification link. Do not replace this OBS route with the maturity procedure of another faculty simply because the credit value is also zero.

35. Do not assume the maturity-test language from the programme language alone

Although the Business Analytics degree is taught in English, maturity-test language also depends on the student’s previous education and statutory language requirements. Follow the current OBS and University rules for the individual case. If the correct language is unclear, resolve it with Oulu Business School before booking the exam. A guide should not tell every international student to use English automatically when the University explicitly links maturity language to prior educational background.

36. Laturi is the formal thesis workflow and archive route

Laturi is not only a place to upload the final file. The University describes it as the system for launching, supervising, monitoring, evaluating and publishing the thesis process. After approval, the final thesis is transferred to the University archive and, if the student permits, published in OuluREPO. Treat the final upload as a controlled release: freeze the manuscript, verify title and metadata, remove confidential information, confirm figures and tables against the final analysis and preserve the submitted version separately from working drafts.

37. University rules give thesis examiners one month for statements

Current University assessment rules state that examiners of a Master’s thesis must issue their statements within one month after the final-form thesis is submitted. The publicly accessible OBS material used for this guide does not establish a Business-path-specific examiner count or exact examiner composition. Therefore, plan time for formal examination but verify the assigned reviewers/examiners in the current Laturi and OBS process rather than importing a number from the technical Business Analytics route.

38. Plan graduation separately from thesis completion

For Business Analytics MSc EBA, the current graduation page instructs students to contact Academic Affairs about one month before graduation so that required studies can be checked and the graduate questionnaire can be provided. After that check, the degree-certificate application is submitted electronically through Peppi. The page also provides a current thesis-return and graduation schedule for fall 2026 and spring-summer 2027. A finished manuscript alone does not guarantee the next possible graduation date.

39. A practical 12-step workflow for Business Analytics Business

1. Confirm programme 51732 and your current PSP. 2. Confirm 721020S / 30 ECTS as your thesis object. 3. Enrol in the current thesis implementation and obtain topic approval. 4. Build the research plan around a defined business problem and academic question. 5. Resolve data access, confidentiality, personal-data and ethics requirements before processing. 6. Complete the research-plan seminar stage. 7. Build theory and method into the intermediate report. 8. Preserve data provenance, code and analytical decisions while conducting the study. 9. Complete the manuscript and oral presentation. 10. Perform the required opponent role and seminar participation. 11. Complete 721010S and submit the public-safe final thesis through Laturi. 12. Complete Academic Affairs checking and apply for the degree through Peppi.

40. Final pre-submission checklist

Verify that 721020S remains the current 30 ECTS thesis, the seminar is still embedded in that course, and 721010S remains the required 0 ECTS maturity test. Confirm supervisor approval, required seminar/report/opponent activities, current evaluation rubric, data and ethics permissions, public-safe company material, reproducible analysis, correct unit of analysis, appropriate validation, limitations, maturity-test arrangement, Laturi submission and the current Business Analytics MSc EBA graduation timetable. Finally, recheck the curriculum boundary: the University is moving to the 2027-2030 curriculum regime, so current 2026-2027 details should be freshly verified before they are treated as final rules for autumn-2027 study arrangements.