Choice-based conjoint (CBC) analysis reveals how customers trade off product attributes when making purchase decisions. Each respondent completes multiple choice tasks where they select their preferred alternative from a set of products with varying attribute levels (brand, features, price, etc.).
The tool estimates individual-level part-worth utilities using multinomial logit (MNL) regression. For each respondent \(i\), the utility of alternative \(j\) in task \(t\) is:
$$
U_{ijt} = \alpha_{j,i} + \sum_{a \in \text{cat}} \sum_{\ell} \beta_{a,\ell,i} \cdot I(\text{level}(a) = \ell) + \sum_{b \in \text{numeric}} \beta_{b,i} \cdot x_b + \beta_{\text{price},i} \cdot p_j
$$
where \(\alpha_{j,i}\) are alternative-specific constants (for competitor and "None" alternatives), \(\beta\) coefficients represent part-worth utilities for each attribute level, and \(\beta_{\text{price},i}\) captures price sensitivity.
The probability respondent \(i\) chooses alternative \(j\) follows the multinomial logit:
$$
P(y_{it} = j) = \frac{\exp(U_{ijt})}{\sum_{k \in C_{it}} \exp(U_{ikt})}
$$
Key Concepts
- Part-worth utilities: The value (in utility units) a customer assigns to each attribute level. Higher values = stronger preference.
- Attribute importance: The range of utilities within an attribute divided by the sum of all ranges. Shows which attributes drive choice most.
- Price coefficient: Usually negative; larger magnitude = more price-sensitive. Used for willingness-to-pay calculations.
- Individual-level estimation: Each respondent gets their own coefficients, enabling segmentation and personalized simulations.
- None alternative: Models the "no purchase" option with its own constant term per respondent.
- Competitor alternatives: Real products with known attribute profiles, modeled via alternative-specific constants.
Understanding the "None" Option (Opt-Out Modeling)
How it works: The "None" alternative is modeled as a constant utility term (ASC_None) with no product attributes. This represents a respondent's baseline preference for "doing nothing" or "not purchasing."
Why constant-only? Since "None" has no tangible product features (no brand, no price, no screen size, etc.), its utility doesn't vary with attribute changes. It serves as a fixed outside option that respondents can always choose instead of any product.
Interpretation:
- Negative ASC_None (e.g., -2.5): Respondents generally prefer buying something over nothing. The more negative, the stronger the "must have a product" preference.
- Positive ASC_None (e.g., +1.2): Respondents prefer doing nothing unless a product offers substantial utility. Indicates high barriers to purchase.
- ASC_None ≈ 0: Indifference between "None" and a product with baseline attribute levels. Market is highly competitive for attention.
In market simulation: When you change product configurations (e.g., lower price, better brand), the "None" utility stays constant while product utilities change. This allows realistic "no purchase" share predictions that reflect actual market opt-out rates.
When to Use Conjoint Analysis
Use conjoint when you need to understand customer preferences for multi-attribute products or services. Common marketing applications:
- Product design: Which features drive purchase intent? What's the optimal feature set?
- Pricing strategy: How much can we charge for premium features? What's willingness-to-pay?
- Market simulation: If we launch Product A vs Product B, what market share will each capture?
- Competitive positioning: How do customers value our brand vs competitors?
- Segmentation: Do different customer segments value attributes differently?