Unlocking hidden market segments: A data-driven approach exemplified by the electric vehicle market

Abstract

Market segmentation is crucial for companies to recognise the distribution of products in the market and to identify ‘unexploited’ segments that hold the potential for new products not yet available in the market. However, recognising market segments that are not yet occupied by any product requires extensive research and data analysis. To address this challenge, we present a new systematic, data-driven approach to market segmentation based on product attributes data. This approach combines three data mining methods (singular value decomposition, principal component analysis, and clustering) with a newly developed inverse clustering algorithm. Inverse clustering introduces interpretable variables (i.e., principal components) and quantitatively identifies unexploited market segments distinct from existing ones. We apply this approach to a use case of battery electric vehicles to demonstrate its effectiveness in supporting product positioning and analysing market data. Leveraging the developed techniques and algorithms could bridge the gap between product development and market potential by identifying opportunities for new products. The approach offers better explainability and applicability of market segments, effectively identifying unexploited market segments that traditional market research methods may have overlooked.