How to do Market Segmentation

Market Segmentation is the cornerstone of most marketing strategies. This ebook will show you how to do it well so you set your clients up for success. Whether you’re creating your own segmentation for the first time, or you want to hone the approach you already use, this ebook is for you. From writing "golden questions" to choosing the right algorithm to identify segments, this book will show you many ways to improve your market segmentation process.

What you'll get from this ebook

This book covers the core steps you need to perform when doing your own segmentation:

  • Writing ‘golden questions’ for market segmentation
  • Choosing the right algorithm (cluster vs latent class analysis)
  • Treatment of missing data
  • Standardizing data
  • How to trade-off statistical versus managerial considerations when choosing your final segmentation

More about this ebook

The How to Do Market Segmentation ebook provides a detailed set of instructions explaining how to create segments using survey data. A basic familiarity with surveys and market segmentation is assumed.

Data preparation
1. Don’t knows and non-responses
2. Missing data due to not being asked (skips and filters)
3. Ordinal Variables
4. Nominal Variables
5. Nominal variables have more than 3 categories
6. Questions with different ranges
7. No or few observations have complete data
8. Missing data can be predicted based on non-missing data
9. Data is missing for a really good reason (nonignorable)
10. Weights
11. MaxDiff, Conjoint, Driver Analysis, and Choice Data

Forming segments
12. Segments that mainly reflect response biases
13. Weird results from latent class regression
14. Visualizing differences between segments
15. Segments differ on only a few variables
16. Identifying “golden questions”
17. Working out the best number of segments
18. Algorithm has not converged
19. Local optima
20. Recreating segments in a new data set
21. Measuring replicability

How to do market segmentation
22. Increasing replicability
23. Quickly profiling segments
24. Increasing the predictability of the segments
25. Finding the most managerially useful segments

Read more

Get started today