What needed to be understood.
Customers evaluate a product as a bundle of attributes. The analysis needed to separate those combined preferences into interpretable utility values and compare realistic configurations.
How the analysis was built.
- Encoded attribute levels for regression-based conjoint estimation.
- Estimated part-worth utilities and interpreted positive and negative preference signals.
- Converted utility differences into approximate dollar values where supported by the model.
- Compared alternative product configurations and checked design orthogonality.
What the evidence showed.
- Attribute levels contributed unevenly to total preference, revealing which features drove product appeal.
- Utility-based comparisons made it possible to distinguish a best-value configuration from a more premium option.
- Orthogonality checks helped evaluate whether the experimental design could isolate attribute effects.
What should happen next.
- Prioritize features with strong utility contribution rather than adding costly attributes indiscriminately.
- Validate willingness-to-pay estimates through market tests before using them as final prices.
- Use simulation to compare proposed configurations against likely competitive alternatives.
What this project does—and does not—prove.
Conjoint estimates reflect the supplied choice/rating exercise and model specification; actual purchases can differ from stated preferences.