Market Segmentation and Analysis Techniques

Posted on Jan 24, 2025 in Statistics

Market Segmentation Techniques

Cluster Analysis: This is a segmentation technique, a process by which we identify groups of consumers according to particular characteristics, with the aim of searching for a differentiated offer for each. We seek:

  • Uniformity Inside: Maximize the difference between groups.
  • Heterogeneity Between Groups: Maximize the difference between groups.

Conclusions: To determine what group/name and p-value > 0.05, accept the null hypothesis (H0) which means no difference between groups. Is this useful? If there is a difference, yes. Then, what do we do? We look at the socio-demographic information.

Starting H0: H0: Independent or no association between x and y; H1: x and y are associated. Conclusion: Make a frequency table (frec2); Ee = (Vo-Ve)² / Ve + … compare it to the Chi-square and reject or accept H0.

Association Analysis

Association:

  1. Coefficient PHI = √(Eexp / n)
  2. Coefficient of Contingency = √(Eexp / (Eexp + n))
  3. Coefficient Cramer = √(Eexp / n) / min(r-1)(r-1)

All values < 0.6 indicate a weak association between variables.

Steps to Follow in Cluster Analysis

  1. Selection of variables: Variables to be included in the analysis should be scale variables (containing continuous data).
  2. Election of the average distance between cases: We calculate the Euclidean distance.
  3. Choosing of classification: Two types: Cluster hierarchy (for sample size < 100) and non-hierarchy (> 100).

Timeline for Cluster Analysis

  1. Define the continuous scale variables used in the cluster analysis.
  2. Fix the number of clusters.
  3. Determine those variables where there are significant differences.
  4. Analyze the sociodemographic characteristics of the cluster.
  5. Apply the most suitable strategy for each cluster identified.

Note: The smaller the number of clusters fixed a priori, the more reliable the result in obtaining significant differences between groups.

Experimental Design

What type of experimental design do you use? Why? (It should be in the middle column and row, total and average). The experimental design is the Latin square. It is the best suited to the popular features of it. Characteristics: Two external variables and one treatment variable; test variables are not all homogeneous.

(Make total treatments and averages) D.total = D.treatment (wreck) + D.month + D.establishment + ERROR (the difference).

Draw a table!

  • Rows (x order): Sources of dispersion; Sum of squares (D); Degrees of Freedom; Variance (Sum of squares / Degrees of Freedom); F. Snedecor (Variance / Error) = Result ≥ Result in tables (19).
  • Columns: Price, month, provider, error, total.

If they are >, reject H0; if they are <, accept H0.

The question is: Which variables are determined in the research of “what is it”? Those that reject the H0. “What is the right price?”: If the price influences sales (H0 is rejected), we face inelastic demand, so we shall maximize the price to maximize profit.

Marketing Research Methods

Phase I: Determine the need for a report

  1. Define the problem and opportunity. Address general questions to resolve, decision-maker, and investigate the causes of specific objectives.
  2. Establish the marketing research.

Phase II: Design the marketing research

  1. Choose the type of investigation: Exploratory, descriptive, or causal.
  2. Plan the sampling (define the universe of study, sample size, and sampling procedure).
  3. Establish the information support, scales, and measures (choose, inform staff, and design a postcard questionnaire).

Phase III: Implementation of the marketing research

  1. Collect data.
  2. Analyze the report.
  3. Obtain results.

Phase IV: Communication of results

  1. Prepare the results report (design and preparation).
  2. Present the results: conclusions and recommendations.