top of page
GARRICK TAYLOR BYRNE
RETAIL STORE LOCATION SELECTION
Bicycle Store in Germany
Premise:
I was hired by an entrepreneur who wished to open a chain a bicycle shops in Germany.
There are four criteria that the client believed to be important for choosing a location: population density, mean age, proportion of male residents, and household size. Using census data, I ranked the populated areas of Germany according to these criteria. Click on the maps below for an explanation of each criterion.
Analytical Methods:
For each of the four criterion, I ranked every populated square kilometer in Germany on a scale of 1 to 5, with 5 being the most attractive and 1 being the least. Then, I summed each area's score from each criterion map to calculate a total score.
​
Results and Analysis:
The maximum possible score was 20 (4 criteria x maximum rank of 5). The highest score achieved was 12. One area scored 11, and 69 scored a 9 or 10.
​
As expected, the highest scoring locations were in city centers.
​
Our client was impressed with my analysis, but he forgot to mention something to me. Naturally, the client wanted to avoid opening a store in an area that's already saturated with similar businesses. The client agreed that the stores are best located in metropolitan areas, but he felt that population density is of less importance and wants to remove population density from the analysis.
​
For each 1-square-kilometer cell, the GIS counted the number of existing bike stores. The maximum number of stores in any square kilometer area was found to be 10. The minimum was zero. Each cell was assigned a rank of 1 to 5, with the score inversely proportional to the number of existing bike stores.
​
The population density statistic was removed from the model.
​
As before, there were four criteria, with a maximum possible score of 20.
​
The new model is shown below. Scores range from 5 to 14.
​
The client now has maps showing areas that would be suitable for his store.
​
Tools and data sources:
The statistic rasters were prepared in R and the maps were assembled in QGIS.
​
Population demographics came from the German census. Existing business locations were obtained from Open Street Map.
​
This demonstration was based on an exercise from the book Geocomputation with R by Robin Lovelace, Jakub Nowosad, and Jannes Muenchow.
bottom of page