Table of Contents
A customer segmentation is arguably one of the most strategically useful research projects you’ll commission. Done well, it will direct your marcomms strategy and your product/service development, ensuring your efforts are focused on the most valuable customer segments in the market place. To find out more about the fundamentals of a customer segmentation, read our article here.
There are two key activities involved in a customer segmentation that are pertinent to this article:
- The gathering of a substantial and robust data set. Primarily captured through an extensive questionnaire administered to a large and broad cross section of category users / buyers.
- The application of advanced analytical tools to identify how many segments existing in the marketplace, what sets them apart from each other (i.e. their demographic, behavioural, and attitudinal characteristics) and the relative size of each segment as a percentage of the whole marketplace.
Add to these the development of segment personas and other deliverables including company roll-out, and as one might imagine, the whole journey of commissioning and then delivering a customer segmentation is intensive and time consuming for all parties involved, including you as the client.
The Shortcoming of a Traditional Segmentation
A customer segmentation captures a snapshot in time and as such has an effective half-life. This is because the further in time we move away from the moment of its inception, the less of a true reflection it is of the current customer marketplace. As a result, the useful ‘working life’ of a customer segmentation is between 3 and 5 years – the differential can be affected by how fast moving a particular category is or a significant market/country/global event. The COVID pandemic is a good example of the latter as it changed the way people work and play, and what they choose to spend their money on.
A Future Possibility
To move beyond this shortcoming, we suggest a reappraisal of the ways in which a customer segmentation is set up, executed, and delivered. An approach that harnesses the proliferation of computational power and digital technologies to capture and analyse data in real-time to create living, dynamic segments that don’t suffer from having a relatively limited working life.
Perhaps the best way to explain this is to call out the traditional approach and then compare that with the suggested new approach. We’ll start with data capture.
- the traditional approach:
- design an in-depth questionnaire comprising of circa 45-50 questions
- capture the data over a short, fixed period in time, circa 1-2 weeks
- survey a robust sample size of circa 1,500 category users / buyers
- the new approach;
- design a series of short, quick-to-answer surveys (approx. 4-5 questions) that are individually discrete but pieced together, overall cover the totality of the questioning including in the traditional questionnaire
- use AI-driven imputation to generate a comprehensive data set, drawn from the discrete variables captured in the bite-sized surveys
- apply an ‘always on’ approach to data capture, ensuring a constant feed of real-time information on customer demographics, behaviours, and attitudes
- survey thousands of category users / buyers
- design a series of short, quick-to-answer surveys (approx. 4-5 questions) that are individually discrete but pieced together, overall cover the totality of the questioning including in the traditional questionnaire
And next we’ll look at the application of segmentation analytics and identification of the segments:
- the traditional approach:
- An intensive two week period of data analysis using statistical methods including Factor analysis, Cluster analysis, and Discriminant Analysis to identify the segments in terms of the factors and features that set them apart from the other segments, and the size of the market each segment occupies.
- Followed by a round of collaborative sessions between the research agency, the analytics experts, and the end client to look over and decide on which ‘segment solution’ to adopt, i.e. do we opt for the four, five, or six segment solution?
- Bring the segments to life as customer personas, presented back in a formal presentation and/or workshop.
- the new approach:
- Retain the period of data analysis using statistical methods including Factor analysis, Cluster analysis, and Discriminant Analysis to identify the segments
- And as before, decide on which ‘segment solution’ to adopt
- Bring the segments to life as customer personas, hosted in an online environment, providing you with constant and ready access to your key customers segments along with the insights into how to most effectively reach and communicate with them
- Then refresh the segmentation analytics every 6-8 weeks to track shifts in behaviours, attitudes, and demographic profiling across and within the segments
Conclusion
By harnessing the current computer power coupled with digital technologies (and untethering ourselves from the traditions of how a customer segmentation is set-up, executed, and delivered) we can move beyond the current shortcomings and provide you (our clients) with a living, breathing reflection of the real-world customer landscape. One that can tell you who your most important segments are, how to connect with them, and how they’re changing over time.
