Work

Questions I've explored

Each project explores a behavioural or business question through data, research, and analysis.

Analysing User Behaviour
How do clients navigate the platform, and which barriers prevent them from finding the information they need?
I analysed clickstream data from De Innovatiespotter’s platform to understand how clients used its features and search functions. The analysis revealed frequent misuse and underuse of key features, and the findings contributed to a platform redesign focused on improving usability and the overall user experience.
RStudioExcelCanvaMachine learningBig data analyticsQuantitative researchQualitative researchUX researchPredictive modellingClustering
Read case study
Predicting Student Dropout
Can machine learning detect dropout risk in online learning?
I conducted my MSc thesis with Abe Hofman and Annie Johansson at Prowise Learn, an online learning platform. I applied machine-learning and statistical models to identify students at risk and predict dropout.
RStudioSQLCanvaMachine learningPredictive modellingLogit leaf modellingMulti-state survival analysisSurvival analysisBig data analyticsQuantitative researchPsychometrics
Read case study
Understanding Teacher and Student Engagement
Does teacher engagement with an online learning platform predict higher student engagement?
I analysed teacher and student activity data to examine how teacher engagement was associated with student usage and progress. Teacher login frequency explained little variation in student outcomes, while strong same-day synchronisation suggested that teachers may use the platform mainly to monitor activity rather than actively shape it.
RStudioCanvaBig data analyticsQuantitative researchData visualisation
Read case study
Repositioning De Bijenkorf for Gen Z
How can De Bijenkorf remain relevant to Gen Z while protecting its premium brand positioning?
I analysed customer reviews, market trends and Gen Z preferences to size the gap between De Bijenkorf's luxury positioning and younger consumers' expectations. We proposed a strategy built on sustainable premium offerings, local brands and experience-led stores, weighed against its commercial and brand risks.
CanvaConsumer researchMarket analysis
Read case study
Designing a Behaviourally Informed HelloFresh Experience
How can consumer psychology be used to improve engagement in the HelloFresh app?
I analysed HelloFresh's brand positioning and customer needs to design a personalised “Add Weekly Menu” feature that uses defaults and social proof to simplify meal planning. A supporting launch strategy paired the feature with loyalty, referral and influencer marketing to drive adoption.
CanvaConsumer behaviourQualitative research
Read case study
Building an Innovation Ecosystem Dashboard
How can complex regional innovation data be transformed into an intuitive tool for non-technical decision-makers?
I designed and built an interactive Power BI dashboard representing regional innovation ecosystems, with a strong focus on information hierarchy and visual storytelling. The four-page dashboard enabled clients to filter the data, explore regional patterns and compare two regions directly, making complex information easier to interpret and use in policy and strategic decision-making.
Power BIDashboard designData visualisationInformation designData storytellingUX designInteractive reportingBusiness intelligenceStakeholder communicationBig data analytics
Read case study
Diagnosing Learning Difficulty and Engagement
How can behavioural exercise data be transformed into a fair and interpretable measure of subject difficulty?
I built a standardised difficulty index from error rates, hint usage and give-up rates, using z-scores to compare subjects on a common scale. This surfaced which topics were relatively harder and revealed engagement patterns the client could act on.
RStudioCanvaLearning analyticsQuantitative researchData visualisation
Analysing LinkedIn Performance
Which content characteristics generate the strongest engagement, and who is interacting with the company's LinkedIn posts?
I analysed post-level LinkedIn data, categorising posts by content type and classifying engaged professionals by job role. Posts featuring maps and municipality comparisons performed strongest — a clear, actionable pattern for future content planning.
ExcelRStudioData classificationQuantitative researchData visualisation

Questions I've explored

Analysing User Behaviour

Predicting Student Dropout

Understanding Teacher and Student Engagement

Repositioning De Bijenkorf for Gen Z

Designing a Behaviourally Informed HelloFresh Experience

Building an Innovation Ecosystem Dashboard

Diagnosing Learning Difficulty and Engagement

Analysing LinkedIn Performance