Cooperation Company Project: A Journey Through Retail Location Analysis
Master in Applied Data Science / 24. April 2024
  • Teilen

  • 2088

  • 0

  • Drucken
Master in Applied Data Science Class of 2024
Rhushikesh is currently in his final semester for Master in Applied Data Science programme at Frankfurt School. He holds 3 years of working experience in IT Consulting and Finance.


Mehr Blog Posts
Turning Ideas into Impact: The Case of Ceres FieldCheck
Erforschung der Auswirkungen von KI auf die Zukunft der Arbeit
Balancing work and study as a Master in Applied Data Science student

In the fast-evolving domain of retail, the adage „location, locality, area“ has never lost its significance. However, determining the optimal spot for a new retail outlet involves a complex interplay of data analysis, artificial intelligence, and a deep understanding of urban geography. My journey along with Ian, Pei and Jinjia through an academic project titled „Finding Optimal Retail Locations“ provided us with an invaluable perspective on these challenges and the innovative methodologies required to address them.

We were introduced to this project in September 2023 as part of a practical project integrated into our academic curriculum. Collaborating with a corporate partner, our group of four students boarded on a journey into the data science landscape. Tasked with applying the knowledge we’d gathered in our previous semesters, our project focused on a tangible challenge within the retail sector: finding the best locations for retail outlets in an urban setting, along with the use of new tools and libraries. To tackle this issue, we utilised a blend of data management, machine learning, and visualisation techniques, diving deep into the complexities of retail location optimisation.

The Framework and Objective of the Project

We used the OpenStreetMap (OSM) platform for our project because it has a lot of detailed maps from all over the world. We decided to focus on Frankfurt, Germany. Our job was to look at different places like restaurants, cafes, and bars and see how close they were to important city spots like bus stops, train stations, and parking areas. We planned to give scores to these things to figure out which areas in Frankfurt were the best for opening new retail stores.

The Challenge of Data

One of the initial hurdles was navigating through the vast, albeit sometimes incomplete, dataset from OSM (OpenStreetMap). With over 1265 restaurants and 217 features examined, the task of cleaning and preparing the data for analysis was daunting. Despite these obstacles, the project allowed us to develop a profound understanding of data acquisition, visualisation, and the intricacies of handling open-source geographical information.

Innovative Methodologies and Results

Our methodology incorporated the use of OSMnx and GeoPandas, tools that facilitated the exploration and analysis of geospatial data. Through a meticulous process of evaluating amenities density and accessibility, we identified Frankfurt Hauptbahnhof and its surrounding area as a prime location, distinguished by its high connectivity and proximity to essential urban facilities.

Encountering and Overcoming Difficulties

The project cannot be completed without overcoming its challenges. The inconsistencies in data formatting and the lack of detailed information for predictive modelling highlighted the limitations of relying solely on open-source data. Furthermore, the absence of defined metrics for optimal location evaluation compelled us to innovate and develop our criteria, focusing on factors such as foot traffic and visibility.

Learning and Growth

This project was a profound learning experience. It honed our skills in exploring geospatial data, visualising complex datasets, and adapting to unforeseen problems. Collaborating closely with our peers and engaging with corporate partners provided us with a glimpse into the consultative process and the dynamics of project management and consulting.

Looking Ahead

The project’s conclusion marked the beginning of a new journey. Our next steps involve extending our analysis through clustering, market segmentation, competitive analysis and finding out new ways to deal with the unknown data. The future scope of this project can potentially be expanding our project to more cities and types of places, and eventually create an interactive tool that could change in innovative ways and can consult retailer to decide where to open new stores.

Through this project, we didn’t just use what we learned in the master’s programme to solve a real problem, but we also learned a lot of new skills. This experience showed us how important it is to work with people from different fields, how powerful data can be in making decisions, and how technology and business can work together in so many ways. As we keep going, the things we learned and the methods we came up with will definitely help us in future projects related to data science and planning for retail businesses.

0 Kommentare