FACTOR, PRINCIPAL COMPONENT AND CLUSTER ANALYSIS
PRINCIPAL COMPONENT AND CLUSTER ANALYSIS IN SEGMENTATION
Data analysis conducted using SPSS and MPLUS
Consumer Insights relating to the choiced of switching energy deal from the Utility company.
The Goal was to produce a bottom-up segmentation (i.e. Data-Driven) of Swiss households in relation to electricity, heating and mobility behaviors.
First, a principal component analysis (PCA) was conducted on a set of 25 original variables, representing three dimensions (equipment, usage, behavior) in three energy fields (electricity, heating, mobility). Second, a cluster analysis was conducted on the principal component scores obtained from the PCA. The final objective of this analysis was to provide a (bottom-up) segmentation of households.
These analyses were written up in an official document and made available to policy-makers, practitioners and stakeholders in Switzerland
Weber, Burger, Farsi, Martinez-Cruz, Puntiroli, Schubert & Volland (2017)
Other similar analyses and interesting documents on the same topic can be found through the link below:
CLUSTER ANALYSIS IN SEGMENTATION
Data analysis conducted using SPSS
Multiple variables were entered into a Cluster Analysis to observe which group of variables best predicted that customers would choose a more expensive energy package from their Utility company. The analyses were repeated many times, either forcing a specific number of clusters in the output or by removing variables that negatively impacted the analysis. Achieved Cluster Quality wasn't special, but FAIR.
These analysese were conducted for SCCER CREST, with the intention of understanding those customers who opt for dearer or cheaper energy deals compared to the ones they already have.