Projects
Participants of the Master of Science in Data Science for Business are asked to carry out a real-life project in small groups in cooperation with a particular company. The purpose is to solve a problem in the field of business (using Data Analytics methodologies). No courses are provided from April to June, enabling participants to work full time on the project. The integrating project offers participants the opportunity to demonstrate their newly acquired skills and competences. At the end of June, participants will provide teaching staff and company-internal supervisors a presentation of the project.
Typically, teams of two students are formed. Under the expert guidance of a doctoral student as well as a professor, a project is carried out. Companies are charged 20.000 EUR + VAT.
For reasons of confidentiality, it is clear that we cannot disclose many details about the content or results of the projects on a public website.
Please contact Prof. Dr. Dirk Van den Poel (mailto:dirk.vandenpoel A T UGent.be) for more information on the projects.
Academic Year: September 2023 - June 2024
Delhaize
LLM (large language model) project
WEngage
LLM (large language model) project: We contributed to the development of WEngage's WEmail commercial offering. The main objective is to augment the quality of customer interactions with the help of AI (by using large-language models). Please continue reading here (a CustomerFirst article) for more details.
Mediahuis
LLM (large language model) project
Ads and Data
Analytics project
Ontex
Analytics project
3 more confidential projects
Analytics and/or LLM projects
Academic Year: September 2022 - June 2023
Engie
Analytics project
WEngage
Predicting Call Center Volume (featured in an article on Tijd.be)
Ypto (IT provider to NMBS/SNCB)
Analytics project
1 more confidential project
Analytics project
Academic Year: September 2021 - June 2022
AG Insurance
Analytics project
Airties
Analytics project
Mediahuis
Analytics project
Sappi
Analytics project
2 more confidential projects
Analytics project
Academic Year: September 2020 - June 2021
Airties
Analytics project
Delhaize nv
Analytics project
Service provider to the Pharma Industry
Analytics project
5 more confidential projects
Analytics project
Academic Year: September 2019 - June 2020
Delhaize nv
Analytics project
Curium Pharma nv
Analytics project
Van Marcke nv
Analytics project
Westtoer nv
Analytics project
Other confidential projects
Academic Year: September 2018 - June 2019
Algemene Pharmaceutische Bond
Analytics project
Easyfairs International nv
Analytics project
Essent.be nv
Analytics project
Mediahuis nv
Analytics project
Other confidential projects
Academic Year: September 2017 - June 2018
NEW: See our blog post for more details.
AG Insurance
Analytics project
Modanisa.com
aCRM project
SMS Timing
Analytics project
AB InBev
Customer acquisition
Other confidential projects
Academic Year: September 2016 - June 2017
See our blog post for more details.
Barco
Analytics project
Sony Music Belgium
Analytics project
Academic Year: October 2015 - June 2016
See our blog post for more details.
ENGIE (formerly known as Electrabel or GDF-Suez)
Analytics project
Lapperre Hearing Systems
Renewal Analytics project
SMS Timing
Analytics project
Total
IoT/Big data project
Wolters Kluwer
Analytics project
Academic Year: October 2014 - June 2015
See our blog post for more details.
Brady
Market research project
Distriplus
Analytics project
ENGIE (formerly known as GDF-Suez)
Analytics project
SMS Timing
Analytics project
Total
IoT/Big data project
Wolters Kluwer Belgium
Analytics project
Wolters Kluwer UK
Analytics project
Academic Year: October 2013 - June 2014
See our blog post for more details.
Alpro
Market research project
Brady
Market research project
Essent.be (http://www.essent.be)
Marketing modeling project
GDF-Suez / Electrabel
Marketing modeling / big data project
Kazou
Marketing modeling project
Leclerc
Marketing modeling / big data project
Academic Year: October 2012 - June 2013
Essent.be (http://www.essent.be)
See our blog post for more details.
Accent HR
See our blog post for more details.
SSZ (http://www.ssz.nl/)
SSZ stands for 'Studiecentrum Snacks en Zoetwaren Benelux', the research center for Snacks and Sweets in the Benelux.
See our blog post for more details.
Academic Year: October 2011 - June 2012
Natuurpunt (http://www.natuurpunt.be)
See our blog post for more details.
DSC
The project was confidential.
SSZ (http://www.ssz.nl/)
SSZ stands for 'Studiecentrum Snacks en Zoetwaren Benelux', the research center for Snacks and Sweets in the Benelux.
Academic Year: October 2010 - June 2011
USG People
Classified.
SSZ (http://www.ssz.nl/)
SSZ stands for 'Studiecentrum Snacks en Zoetwaren Benelux', the research center for Snacks and Sweets in the Benelux.
Academic Year: October 2009 - June 2010
E-commerce company
A leading e-commerce company, that prefers to remain anonymous.
Objectives of the project: Developing a cross-sell model for online purchases.
Data: A rich data warehouse of their website.
SSZ (http://www.ssz.nl/)
SSZ stands for 'Studiecentrum Snacks en Zoetwaren Benelux', the research center for Snacks and Sweets in the Benelux.
Academic Year: October 2008 - June 2009
IJSBOERKE - Belgian Icecream Group (http://www.ijsboerke.be/)
IJSBOERKE nv is the leading premium icecream brand in Belgium.
Objectives of the project:
Data: A rich data warehouse of their home-vending operation.
SSZ (http://www.ssz.nl/)
SSZ stands for 'Studiecentrum Snacks en Zoetwaren Benelux', the research center for Snacks and Sweets in the Benelux.
Academic Year: October 2007 - June 2008
2 projects for DunnHumby (http://www.dunnhumby.com)
DunnHumby is a specialized provider of database management and analytical services active in many European countries as well as the U.S.
Objectives of the project: Due to confidentiality agreements, we cannot disclose the exact content of the projects. 4 MMA students performed two separate projects in teams of two students.
Data: A rich data warehouse of a large retailer.
SSZ (http://www.ssz.nl/)
SSZ stands for 'Studiecentrum Snacks en Zoetwaren Benelux', the research center for Snacks and Sweets in the Benelux.
Tiense Suiker (http://www.tiensesuiker.be)
Academic Year: October 2006 - June 2007
3 projects for DunnHumby (http://www.dunnhumby.com)
DunnHumby is a specialized provider of database management and analytical services active in many European countries as well as the U.S.
Objectives of the project: Due to confidentiality agreements, we cannot disclose the exact content of the projects. 4 MMA students performed two separate projects in teams of two students.
Data: A rich data warehouse of a large retailer.
3M (http://www.3M.be/)
Corona Direct (http://www.coronadirect.be)
Corona is a direct writer (insurance company) operating in Belgium.
Objectives of the project: Data acquisition case for their car insurance based on actual distance driven.
Data: The complete past mailing history of prospects/customers, as well as the purchasing history was analyzed.
Academic Year: October 2005 - June 2006
Kinepolis (http://www.kinepolis.be)
Kinepolis is the largest operator of cineplexes in Belgium with operations in many European countries.
The main objective of this project was to make a prediction of the number of visitors per movie showing (i.e., per movie, per day/session, per complex). Results revealed that good predictive accuracy could be achieved at the session level. At the complex level, predictions were highly accurate.
A rich data warehouse with movie-attendance data was available in Microsoft SQL Server 2005.
Concentra group - Metro (http://www.metrotime.be/)
Metro is a newspaper that is freely distributed mainly in train and metro stations, but also in universities, Panos shops and more than 200 companies. Every day, Metro reaches about 616.800 people which makes it the third largest newspaper in Belgium, and its share still keeps on growing. The website of Metro, Metrotime.be, offers a full and freely accessible news summary complemented with news archives and a fun & games area.
This research project investigated following three questions: 1) What is the competitive position of Metro in comparison with its direct and indirect competitors?, 2) Do consumers perceive a match between Metro and brands from other sectors?, 3) Do consumers perceive Metrotime as intended?
Both qualitative research via focus groups as quantitative research via questionnaires were applied.
Academic Year: October 2004 - June 2005
Landbouwkrediet (http://www.landbouwkrediet.be)
Landbouwkrediet / Credit Agricole is Belgian bank. Farmers used to be the largest part of their customer base. This has changed gradually by organic growth, but also by two recent acquisitions: Europabank and Keytrade Bank.
Two projects were completed for Landbouwkrediet: (1) Churn prediction: Building a general churn model; (2) Cross-sell: Given the focus on mortgage customers, MMA students built a cross-sell model for mortgage customers.
Data: A rich data warehouse with customer information was available in Oracle
Sint-Blasius (http://www.azsintblasius.be)
Objectives of the project: Branding used to be limited to profit organizations. However, more and more non-profit organizations are starting to pay attention to their (brand) image. Hospitals are good examples of such organizations. This project comprised study of the image of the hospital 'Sint-Blasius'. By means of focus groups and a questionnaire
Tiense Suiker (http://www.tiensesuiker.be)
Objectives of the project: Research on the attitude towards sugar in Belgium and the effect of entrance of private labels on the brand Tiense Suiker: The study consists of two parts. On the one hand, the attitude of Flemish consumers towards sugar will be studied. Qualitative research will be applied for this part. On the other hand, the effect of the liberation of the sugar market in Europe in 2006 on the brand Tiense Suiker, will be investigated. To achieve this, the different factors that have an important influence in the purchase of sugar will be determined. For this second part, quantitative techniques will be used.
Academic Year: October 2003 - June 2004
IJSBOERKE (http://www.ijsboerke.be)
IJSBOERKE is (among other activities) a home vending company of ice cream. They have more than hundred salesreps visiting their customers all over Belgium following a regular schedule.
They asked us to investigate ways to increase the hit rate (i.e., number of successful visits over total number of visits). This is done by predictive modeling, i.e., predicting the probability of a purchase. Moreover, this project consists of a field test of other direct marketing methods to reduce their current churn rate.
Data: A rich data warehouse with call & sales data was available in Oracle
Vanden Borre (http://www.vandenborre.be)
Vanden Borre is a home-appliance retailer selling TV's, VCR's, dish washers, friges, ... .
They want to gain more insight into the choice process of store format in the sales of household appliances. They have an extensive database of previous purchases. We will estimate a multiattribute choice model (similar to the ABB case, see Gensch et al. (1990) in Interfaces) investigating the choice of store format (they are competing against local independent stores and against some other retail chains). This will be done by collecting questionnaire data and combining them with data warehouse data.
Utopolis (http://www.utopolis.be)
Utopolis operates a group of cinemas in middle-sized cities in Belgium, Luxemburg, the Netherlands and France.
Objectives of the project: New digital technology offers opportunities for new products to offer in a cinema. Not only will it dramatically impact the way of doing business as a movie theatre, new contents besides movies could be offered in the future, both real time and recorded material. Utopolis wants to analyze the market potential for these new products. A qualitative research will identify the attitude of specific target groups towards new contents offered in a cinema, and the best way to reach the target audience.
Scotts (http://www.scotts.be)
Scotts is a company specialized in the production and distribution of fertilizers, potting soils, herbicides, weed killers and fungicides for the gardening sector.
The purpose of the project is to answer Scotts' management questions, which are related to three topics: 1) insight into the gardener's profile; 2) user profiles of several product categories; 3) attitudes towards compost and pesticides. To answer these questions, both a qualitative (individual interviews) and quantitative study (800 questionnaires) will be done in Flanders, Wallonia and Holland. Each time, active and passive gardeners (=people who work less than four hours a week in the garden) will be tackled.
Tiense Suiker (http://www.tiensesuiker.be)
Objectives of the project: Research on several pricing issues in the Belgian sugar retail market: the study consists of two parts. On the one hand, the fundamental meaning of sugar for consumers will be studied. Qualitative research will be applied for this part. On the other hand, we will try to give an answer on several pricing questions of the company Tiense Suiker. For this second part, quantitative techniques such as adaptive conjoint analysis and other price-acceptation tests will be used.
Academic Year: October 2002 - June 2003
Canal+ (http://www.canalplus.be)
Canal+ is the largest European Pay TV operator.
Objectives of the project: Churn prediction: many subscription businesses (in this case the Pay TV business, other examples of subscription businesses include magazine subscriptions, mobile phone subscriptions, ...), are facing important challenges to retain their customers. The Pay TV business in particular competes with free-TV, a visit to the cinema, buying DVDs,... . Therefore customer relationship management needs to focus its attention on those customers who are most likely to cancel their subscription (i.e., churn). This project builds a customer attrition (churn) score taking into acount as much information as possible.
Data: A rich data warehouse with subscription data was available in Oracle
Fortis (http://www.fortis.be)
Objectives of the project: Enhancing customer attraction in the domain 'Savings & Investing': The objective of the project is enhancing customer attraction in a significant way in the domain 'Savings & Investing', taking into account the diversity in terms of the needs and expectations of the Fortis customer and the diversity in terms of profit potential. First, a qualitative part contains an in-depth research into the basic motivations for savings and investing, in order to segment the market, position the players and identify the levers to enhance attractivity towards the different segments. The second part of the project consists of the analysis of the database of an existing survey (within the domain 'Savings and Investing'), in order to identify threats and opportunities and make some recommendations regarding the different client types (as defined with Fortis) and the different product groups. A third and final part of the project consists of the review of the questionnaire of the recurrent research (cf. quantitative part), based on the findings of the qualitative research, in order to improve the effectiveness of it.
Creyf's (http://www.creyfs.be)
Objectives of the project: Testing the launch of a new campaign: Creyf's wants to change its positioning in the interim market. Therefore a new campaign is launched of which the objective is to become top-of-mind and to create talk value. Since the front office workers of Creyf's are supposed to be the 'first fans' of the brand, it is important to know what they know and feel about this shift and whether they act upon it. The media campaign is just a 'door opener', but it is the employees in the field who should make it come alive. Therefor the objective of the student project is to find out whether the campaign has an effect on the employees in the field. Three aspects will be dealt with: the know aspect (do they know the campaign), the feel aspect (how do they feel about it) and the act aspect (do they act upon the campaign).
Scotts (http://www.scotts.be)
Scotts is a company specialized in the production and distribution of fertilizers, potting soils, herbicides, weed killers and fungicides for the gardening sector.
The purpose of the project is to answer Scotts' management questions, which are related to five topics: 1) Brand awareness and brand penetration for each product category; 2) Stretch-ability of the umbrella brand Substral; 3) Effectiveness of the commercials; 4) Attitudes towards fertilizers. To answer these questions, both a qualitative (focus group sessions) and quantitative study (800 questionnaires) will be done in Flanders, Wallonia and Holland. Each time, active and passive gardeners (=people who work less than four hours a week in the garden) will be tackled.
Tiense Suiker (http://www.tiensesuiker.be)
Objectives of the project: Research on the attitude towards sugar in Belgium and the effect of entrance of private labels on the brand Tiense Suiker: The study consists of two parts. On the one hand, the attitude of Flemish consumers towards sugar will be studied. Qualitative research will be applied for this part. On the other hand, the effect of the liberation of the sugar market in Europe in 2006 on the brand Tiense Suiker, will be investigated. To achieve this, the different factors that have an important influence in the purchase of sugar will be determined. For this second part, quantitative techniques will be used.
Academic Year: October 2001 - June 2002
AXA Bank Belgium (http://www.axa-bank.be)
The merger of Anhyp and Ippa created one of the largest banks in Belgium. This, together with the insurance-part of former Royal Belge creates one of the most powerful financial groups in the heart of Europe with worldwide operations.
Objectives of the project: Continuation of cross-selling of products: customer relationship management (analytical CRM) by means of a prediction of the next (most likely) product to buy.
Data: The merger of three groups (Anhyp, Ippa and Royale Belge) created one of richest data warehouses in Belgium.
Makro (http://www.makro.be)
Makro is a hypermarket chain operating 6 stores in Belgium. They offer a wide assortment of both food and non-food products.
Objectives of the project: Improving the targeting of the brochure. They send a bi-weekly mail to 650.000 households.
Data: We use the data of their data warehouse containing all purchases for the last 3 years in an Oracle environment
Corona Direct (http://www.coronadirect.be)
Corona is a direct writer (insurance company) operating in Belgium.
Objectives of the project: Improving multi-mailing targeting of customers. Implementing the DIMAROPT model (Direct Marketing Optimization) developed by Prof. dr. Dirk Van den Poel.
Data: The complete past mailing history of prospects/customers, as well as the purchasing history will be analyzed.
HUBO (http://www.hubo.be)
Hubo is the third largest Belgian DIY (Do-it-yourself) retail chain.
Objectives of the project: Finetuning of last year's 'Market Basket Analysis' to find product (-category) complementarities at different levels in the hierarchy. This should lead to an improved composition of the brochure.
Data: In the project, the complete dataset of customer transactions of all retail locations (more than 13 Gb of data) is used.
Neckermann Postorder nv (http://www.neck.be)
Neckermann mail-order (not the travel branch) is the second largest mail-order company in Belgium.
Objectives of the project: We investigate the assortment customers purchase.
Data: Purchasing history for the past 5 years of a large set of customers.
Academic Year: October 2000 - June 2001
3 projects at AXA Bank Belgium (http://www.axa-bank.be)
The merger of Anhyp and Ippa created one of the largest banks in Belgium. This, together with the insurance-part of former Royal Belge creates one of the most powerful financial groups in the heart of Europe with worldwide operations.
Objectives of the project: 1. Customer relationship management (analytical CRM) by means of a prediction of the next (most likely) product to buy. 2. Churn analysis (who leaves the bank, how can the company prevent this from happening, what are key signals,...). 3. Setup of experiments to promote products from different product families. Including cross-sell, up-sell, ... .
Data: The merger of three groups (Anhyp, Ippa and Royale Belge) created one of richest data warehouses in Belgium.
DMD (http://www.zappybaby.be)
Direct Mail Distributors is a sister company of Neckermann Postorders nv (Belgium). It is active in selling niche catalogs. They have operations in Belgium and The Netherlands.
Objectives of the project: Qualitative study about the consumer (mail-order) buying process of one of their catalogs (Baby Walz).
Data: Students use focus group interviews, and personal interviews to carry out exploratory research.
Corona Direct (http://www.coronadirect.be)
Corona is a direct writer (insurance company) operating in Belgium.
Objectives of the project: Database Marketing project on response modeling to improve mailing efficiency.
Data: The complete past mailing history of prospects/customers will be analyzed.
HUBO (http://www.hubo.be)
Hubo is the third largest Belgian DIY (Do-it-yourself) retail chain.
Objectives of the project: 1. Market Basket Analysis: we try to find product (-category) complementarities at different levels in the hierarchy: this part is a continuation of last year's project 2. Calculation of price elasticities, and more in general, the impact on sales (and profitability) of changes in marketing-mix variables.
Delhaize (http://www.delhaize.be)
Delhaize is one of the largest retailers in Belgium. It is the 21st largest retail chain worldwide.
Objectives of the project: 1. Analysis of the loyalty card: segmentation of customers 2. Cross-selling potential of the Delhaize Plus-card 3. Many confidential additional topics
Academic Year: October 1999 - June 2000
HUBO (http://www.hubo.be)
Hubo is the third largest Belgian DIY (Do-it-yourself) retail chain.
Objectives of the project: 1. Calculation of LTV (life-time value) of customer segments: How do customer segments differ in terms of LTV or its components (retention rate & difference between revenues and costs) 2. Market Basket Analysis: we try to find product (-category) complementarities at different levels in the hierarchy
Data: In the project, the complete dataset of customer transactions of all retail locations (more than 10 Gb of data) is used. The BONUSKAART loyalty card makes the LTV calculations possible.
MAXICOLOR (http://www.spector.be)
Maxicolor is the mail-order operation of the Spector group, i.e., they offer photo-finishing services by mail.
Objectives of the project: 1. Calculation of LTV (life-time value) of customer segments. Research question: How do customer segments differ in terms of LTV or its components (retention rate & difference between revenues and costs). 2. Finding a more sophisticated segmentation scheme for mail-order 3. Which of the direct-mail incentives is most effective
Data: As this is a mail-order operation, detailed data is available on the marketing communication towards all customers, as well as all transactional information.