Correlation analysis: Find the relevant data for your location analysis

Correlation analysis in three steps: Our procedure

The WIGeoGIS typically carries out a correlation analysis as a service in three steps: The selection of samples at the locations, calculation of variables describing the environment of all locations, determination of the relationship between the environment and the business success of each location.

Step 1: Defining a representative sample

During the 1st step, we define a representative sample in the existing location network. The following applies: The locations must be as comparable as possible with reference to format, area and location classification. You also have to define one or more key indicators for the location success. This indicator is usually the turnover at the location. It can also be subdivided into various product groups. You can thus determine which location environment is most favourable for the sales of a certain product group.

Step 2: Determination of relevant environmental variables

In the 2nd step, the possible, relevant environmental variables are determined for the sample. This can be calculated on the basis of potential data, which are aggregated in the catchment areas of the locations. Example: How many inhabitants, what purchasing force or what age structure are present within 10 minutes walking distance of the location? Relevant variables can also be provided by the distance from competitors or traffic-generating companies and traffic frequency data at the location. You thus obtain a mix of variables for each location that describes the location and its environment.

Step 3: Comparison of environmental variables with location success

In the 3rd step, we compare the individual environmental variables with the location success. A correlation coefficient is determined for each one. This describes the type and strength of the statistical correlation between a variable and the key indicator. Example: A high positive correlation coefficient variable "Percentage of 20 to 35-year-old inhabitants" within a catchment area provides information that this target group has a very positive effect on the turnover at this location within the branch environment. In addition to the calculation of the correlation coefficients, the analysis results can also be visualised by means of a plot.

Making better purchase decisions for geodata

The results from the geostatistical correlation analysis can therefore be used to provide evidence of which environmental factors are relevant for the evaluation of locations. They provide information about how strongly the environment and the success of the location correlate. The correlation analysis leads to a standardised and statistically-based evaluation of which variables are decisive for geomarketing. This allows you to make a better purchasing decision about geodata products. 

Benefits for location decisions and Category Management

The results of the correlation analysis also form the basis for further analyses. You can make more secure location decisions, for example regarding new openings, closures or relocations. Positive influences of environmental factors can also be used to search in a structured manner for possible white spots on new locations. Negative factors can be used to critically evaluate locations Product-specific correlation analysis also permits you to implement Category Management directly adapted to the location environment.