Week 5 - Conjoint Analysis

Question 1

Conjoint analysis

Answer 1

Measuring customers’ preferences for product features

• identifying consumer’s preferences for different product features by looking at their choices

useful for predicting responses

Question 2

Problems with Preferences Surveys

Answer 2

Low Discrimination - People often say many things are “very important”, hard to see priorities

Social Bias - may not give honest answers

Question 3

Conjoint Analysis Assumptions (2)

Answer 3

1. A product is seen as a collection of different attributes or features.
   
   2. A person’s preference for a product is based on how much they value each of these attributes.

Question 4

trade-off analysis

Answer 4

Consumers decide how they balance or trade-off different product attributes (features) when making a choice.

Question 5

Alterantive Conjoint Methodologies

Answer 5

1. Traditional Conjoint
   - respondents rate profiles with ALL attributes. (every possible combinations)
2. Adaptive/Hybrid Conjoint
   - when there are many attributes
   - computer algorithm customizes the questions based on most important factors, focusing on relevant features and trade-offs.
3. Choice-Based Conjoint
   - Instead of evaluating each attribute separately, respondents are asked to choose their preferred option from the set, mimicking real-world buying situations where they compare products.

Question 6

Rules for choosing attributes and levels

Answer 6

attributes:
-easy to understand
-clearly defined, measured
-too many attributes can decrease realiability
-each attribute should have 2-4 levels
-should not be too similar (to prevent multicollinearity)

Question 6

Unacceptable Concepts + Remedies

Answer 6

Products too simple or obvious
Attributes too unrealistic or impossible

Remedies:
Deleting not recommended - can disrupt
Rather create a new set of P combinations

Question 6

Product Concepts + 2 types + challenges

Answer 6

Combinations of attributes
- zero similarity between attributes
- each level (within an attribute) shown EQUAL number of times (ensures full representation)

1. Full Factorial Design
   -all possible combinations (attributes x levels)
2. Fractional Factorial Design
   -a subset of the product concepts

Challenges:
-designs strive to minimize correlation between attributes and maintain balance

Question 6

Applications of Conjoint Analysis

Answer 6

Demand estimation
Simulation of competition and competitive response
Product line optimization
Segmentation

Question 6

How many Product Concepts?

Answer 6

1. We need to calculate parameters first
   N - n + 1 = parameters

*N = total # levels
*n = total # attributes

aiming for 2-3x the parameters
(The more product concepts, the better!!)

Question 7

Collecting Data Preferences

Answer 7

1. Rating of Individual Concepts
   - each P concept seperately rated (1-5 scale)
2. Ranking Concepts
   - example: set of 3 phones, rate best to worst
   - harder to analyze
3. Paired Comparisons
   - choose between 2 P concepts
   -harder to analyze
4. Choice
   - Choice-Based Conjoint
   - choose their favorite option from a set of P concepts (e.g. 3 phones)
   -the hardest to analyze, but reflects real-world buying scenarios