How has this evolved?
In the past segmentation relied on attitudinal data captured through focus groups or interviews which are then fed into a quantitative survey. Statistical measures help us understand how the attitudes and demographics clump together in segments.
Today the proliferation of customer data means that we can build better segmentations because they’re based on data harvested from a range of sources. (e.g. social media, census, passive tracking of online behaviour) So we’re asking people fewer questions and using real behavioural or transactional data (e.g. purchase or loyalty data) which is more precise than relying on people’s memories or perceptions of what they do, which is notoriously inaccurate.