Customer data being gathered must confirm or challenge a brand’s existing thinking.
Marketers that rely solely on demographic consumer segmentation may have limited or misleading understanding of consumers. The author lays emphasis on letting customer personas emerge from data. Depending on demographically segmented customer personas could lead to sub-standard segmentation.
This will ultimately result in comparatively “mediocre” messaging, campaigns and ads. Instead, businesses are advised to leverage unsupervised learning, a type of machine learning algorithm. Through unsupervised learning, segmentation is data-based and consumers are clustered as homogenously and efficiently as possible.
The resulting clusters contain a mix of demographic variables and can lead to insights about what customers value. Such segmentation helps refine marketing strategies to match the newly discovered market segments.
[4 minute read]
