How to Master Market Segmentation for Smarter Product Strategies

What Is Market Segmentation?

Before discussing market segmentation research, we first define it: Market segmentation was first proposed by American marketing scholar Wendell R. Smith in 1956, dividing the overall market into sub‑markets with common characteristics based on user traits.

Why Do Market Segmentation and What Is Its Significance for Companies?

In a market, user needs are diverse. Companies need differentiated strategies for different groups, and market segmentation provides the premise for such strategies.

Segmentation splits the overall market into distinct sub‑markets, allowing firms to understand product‑consumer fit, offer products or services that meet specific segment needs, avoid sending irrelevant messages, and maximize resource efficiency.

Market Maturity Considerations: When to Segment?

Maslow’s hierarchy of needs describes five levels from physiological to self‑actualization. Lower‑level needs have stronger driving force; higher‑level needs emerge only after lower ones are satisfied.

Similarly, in a product category, consumer needs exist at different levels. As a market matures, competition intensifies, products diversify, and higher‑level needs appear.

For example, early‑stage female skincare users focus on cleansing; as the market matures, they demand moisturizing, brightening, and other advanced benefits.

Only when a market reaches relative maturity do diverse needs emerge, providing conditions for meaningful segmentation.

How to Segment the Market: Classification Criteria

Market segmentation uses user characteristics to create sub‑markets. Common segmentation variables include:

Demographic characteristics

Psychographic characteristics

Geographic characteristics

Needs characteristics

Usage‑behavior characteristics

Variables can be used singly or in combination. Demographic, geographic, and usage‑behavior variables are frequently used because they are readily available in company databases and help quickly identify target groups, though they may not explain underlying demand reasons.

Psychographic and needs variables provide deeper insight into consumer motivations but are harder to obtain and often require qualitative research and quantitative data collection.

Examples of segmentation results using demographic variables and psychographic variables are shown in the following images:

Research Method Overview: How to Conduct Market Segmentation?

Market segmentation involves multiple variables. We typically use powerful canonical correlation analysis combined with cluster analysis.

Canonical correlation analysis studies relationships between two sets of variables. It includes linear (suitable for numeric variables) and nonlinear (applicable to numeric, ordinal, categorical, or mixed variables). Because market research involves diverse variable types, nonlinear canonical correlation is commonly used.

Nonlinear canonical correlation requires integer values, so preprocessing steps include:

For decimals, change measurement units or recode.

If the minimum value is 0, add 1 to all values.

For negatives, subtract the minimum value and add a constant.

If too many variables are entered, factor analysis is used for dimensionality reduction before canonical correlation.

The segmentation process typically follows these steps (illustrated in the diagram below) and may include discriminant analysis for validation:

Practical implementation with SPSS:

Step 1: Perform canonical correlation analysis.

Step 2: Conduct cluster analysis to obtain segmentation results.

Step 3: Use discriminant analysis to verify classification results.

How to Determine the Effectiveness of Market Segmentation?

After obtaining segmentation results statistically, we evaluate effectiveness based on these criteria:

Significant separation: distinct groups differ on key variables.

Identifiable/reachable: the target segment can be clearly defined and accessed.

Sufficient size: the segment is large enough to justify dedicated strategies.

Actionable: the segmentation aligns with business objectives and guides product or marketing actions.

To refine results, we may adjust input variables in canonical correlation or modify the number of output clusters (typically 4‑8 groups).

How to Apply Segmentation Results in Marketing Practice?

Segmentation should drive business strategy, not exist for its own sake. Applications include:

Product strategy: tailor product lines to meet the hierarchical needs of each segment, optimizing existing offerings and guiding new development.

Marketing strategy: design distinct marketing plans for each segment based on their product‑use and behavior characteristics.

Audience strategy: use key variables to quickly identify which segment a user belongs to, enabling targeted acquisition or retention tactics.