Rethinking customer segmentation

Two developments ensure that no stone will be left unturned when segmenting customers in the future: On the one hand, research findings from behavioral psychology are slowly seeping into marketing practice. On the other hand, legal regulations on the use of cookies are changing what was previously data-driven marketing into customer-driven marketing. In customer-driven marketing, gaining as comprehensive an understanding of the customer as possible is of central importance, as this is the only way to create a holistic and personalized customer experience. To date, the development of a comprehensive understanding of the customer, or customer intelligence, has primarily relied on interaction and transaction data. Although this allows segments to be modeled, they are incomplete or not all-encompassing. The needs level is missing. This can be supplemented with findings from behavioral economics and behavioral psychology.

Paradigm shift: dynamic, pointed segmentation

Effective marketing requires segmentation - this is nothing new. But up to now, segmentation has only been carried out on the basis of socio-demographic data, website data and possibly initial generic affinities. Psychographic dimensions or predictions about future user behavior are only sparsely modeled and even less used. It is time for a paradigm shift: away from broad segmentation and towards dynamic and pointed segmentation. The merging of the following analysis disciplines will lead to success:

• Identity Resolution The generation of data protection-compliant, consolidated user identities
• Data Science The statistical analysis and interpretation of existing data and the derivation of patterns and trends for the development of predictive models (predictive analytics)
• Behavioral science The qualitative enrichment of the collected data with psychographic characteristics (especially basic needs and relevant decision-making patterns)

Intelligent segmentation depends on the linking of two data sets. Website data provides information about user behavior in terms of recency, frequency or engagement. Data from the database provides information about customer history (transactions, CLV, returns or customer quality).

These databases are linked via an identifier (usually an e-mail address). This makes it possible to develop intelligent segments.

Behavioral economics in segmentation

Linking the two databases enables the profiles or segments to be enriched with the characteristics that, according to current findings in behavioral economics and behavioral psychology, are responsible for more than 95 percent of all decisions: those at the needs level.

The interdependence between psychology and marketing technology is, bizarrely, as clear as it is unknown.

A well-known example still demonstrates the data dilemma well: Ozzy Osbourne and Prince Charles do not differ in their socio-demographic characteristics (male, British, >70 years old, very wealthy, large country residence, etc.). They can therefore leave comparable interaction profiles in the store. However, they could hardly be more different in terms of personality and needs. A purely data-based approach logic must therefore almost fail.

The needs dimension is therefore the decisive new criterion for the successful design of customer journeys and, above all, one that has hardly been considered to date. The fact that it is not yet on everyone's lips can be explained on the one hand by the fact that research findings are only gradually being put into practice. On the other hand, it is also due to the complexity of tapping into this level, which should not be underestimated.

Two ways of addressing customers on a psychological basis

A psychologically based approach in customer-centric marketing can be realized in two different ways: conceptually and automatically.

1. The conceptual approach is a data-based, empathetic approach. Experts such as a customer journey manager, who has extensive customer knowledge and psychological expertise, add psychographic characteristics to the existing audiences or segments.
2. The automated approach is currently still at an early stage. Innovative start-ups such as Behamics in Switzerland are already demonstrating that machine learning algorithms can be used to deliver psychological triggers very effectively. However, psychological knowledge is also required when designing the triggers and the interaction of the systems must be clarified.

No matter which path is chosen, the human need trigger conversion chain can unfold its effectiveness in the interplay between data and psychology.