Classifying Respondents Into The Right Groups
CFR Solutions uses discriminant analysis to build predictive functions that sort respondents into known groups, such as buyers versus non-buyers or brand switchers versus loyalists, based on the variables that separate them most.
Every Group Separated, Every Case Classified
Discriminant functions are built from the predictor variables that separate your groups most cleanly, validated with a hold-out sample before they're trusted, and every classification comes with a hit-rate so you know how much to rely on it.
Classification Models Matched to Your Groups
From a simple two-group split to a multi-category classification, we build the discriminant model that fits your data and your decision.
Two-Group Discriminant Analysis
Classifies respondents into one of two known groups, such as purchasers versus non-purchasers.
Learn more →Multi-Group Classification
Extends the model to sort respondents across three or more predefined segments at once.
Learn more →Brand Switcher Profiling
Identifies the predictors that separate loyal customers from those likely to switch brands.
Learn more →Classification Accuracy Testing
Hold-out validation and hit-rate reporting so you know how reliable each model is before you use it.
Learn more →Stepwise Variable Selection
Narrows a long predictor list down to the small set that actually drives group separation.
Learn more →Segment Prediction Scoring
Scores new or untested respondents against an existing model to predict which segment they fall into.
Learn more →Explore Our Other Essential Services
Every technique is built on the same rigorous, quota-clean data — pick the analysis that matches the question your study needs to answer.
Brand Mapping
Perceptual maps showing how brands sit relative to each other on key attributes.
Learn more →CHAID Analysis
Decision-tree segmentation that splits respondents by the strongest predictors.
Learn more →Cluster Analysis
Groups respondents into natural segments based on shared attitudes or behavior.
Learn more →Conjoint Analysis
Measures the trade-offs respondents make between product features and price.
Learn more →Correlation Analysis
Measures how strongly two or more variables move together.
Learn more →Discriminant Analysis
Classifies respondents into known groups using their strongest predictors.
Learn more →Factor Analysis
Reduces a long list of variables down to the underlying factors driving them.
Learn more →Multidimensional Scaling
Visual maps of how respondents perceive similarity between items or brands.
Learn more →Regression Analysis
Quantifies how much each predictor moves the outcome you care about.
Learn more →