Segmentation Case Study

Client: Global Insurer with operating profit of over $5.5Bn

Objective: To deliver an attitudinal and life stage segmentation driven marketing strategy

Requirements:

  • The segmentation must be derived from the market, not the customer base
  • The segmentation must be actionable – there must be behavioural, attitudinal and “needs” based insights that can be acted upon to increase marketing effectiveness
  • The segmentation must be able to be applied to internal and external databases for profiling, tagging and selection

Segmentation Case StudyWorking alongside our friendly market research agency, data was collected from 15,000 people with a balanced methodology to remove sampling bias, channel bias, etc… 145 questions were asked on their attitudes to financial products, shopping behaviours and demographic factors.

Principle Component Analysis was then applied to these data to identify a set of 13 principle components (abstract variables which are linear combinations of the full set of 145 variables, which are independent of each other, but account for all of the variability in the data).

K-means clustering was then applied to these 13 factors (the principle components) to define 6 market segments with the following features,

  • They must be meaningful and actionable.  This means that they are significantly distinct from each other, both mathematically and in the context of the product and market
  • They must have significant sizes
  • Must contain enough demographic and product level variation to enable “tagging” onto the client database

Once the segments were defined, they were profiled against the original fieldwork and external databases to create rich and detailed profiles of the segments, including,

  • Demographic: Age, Gender, Social Class, Life stage, Marital Status, type of house ownership, etc…
  • Profitability: Conversion rates, income, profitability, etc…
  • Attitudes: Attitude to investments, shopping channel preference, price vs. product preference, etc…
  • Needs:  Product feature preferences, advice requirements, etc…
  • Media Consumption: attitudes to DM, TV viewing habits, newspaper consumption, etc…

The final step was to allocate segments to the customers in the client’s database.  This was achieved with a mix of Logistic Regression Analysis and Discriminant Analysis.

So, now we understood how the market broke down into significant, distinct and actionable segments, the needs, attitudes, life stage and demographic profile of these segments, and we can find them on their customer database.  Working with the client’s media agencies, it was then possible to design a proposition, advertising strategy, retention strategy and pricing strategy for these segments in order to grow target segments to maximise profit.

Please get in touch with us to discuss your segmentation requirements.

Add Comment Register



Leave a Reply