Making Use Of Conjoint Analysis

Conjoint analysis is a marketing research method used to figure out how people value different features of a product or service.

Instead of just asking people directly (“Do you like this feature?”), conjoint analysis forces trade-offs by presenting choices where several attributes vary at once — just like in real life when you choose between products with different prices, qualities, or designs.

In simple terms:

• You create hypothetical products by combining different levels of key features (like price, colour, brand, etc.).

• You ask participants to choose between sets of these options or rank/rate them.

• From their choices, you can calculate the value (“utility”) they assign to each feature.

---

Example:

Imagine you’re designing a new coffee:

• Features: Roast level (light, medium, dark), Origin (Colombia, Ethiopia, Brazil), Price ($3, $5, $7).

• You might show participants:

  • Coffee A: Medium roast, Ethiopia, $5

  • Coffee B: Dark roast, Brazil, $3

  • Coffee C: Light roast, Colombia, $7

They pick their favorite.
Repeat across different combinations.

From enough choices, conjoint analysis can reconstruct how much people care about:

• Dark vs. light roast

• Origin

• Price sensitivity

---

What does it tell you?

• Which features drive consumer decisions the most (e.g., price more important than origin).

• How much extra people are willing to pay for a certain feature (e.g., $2 more for Ethiopian beans).

• Optimal product designs for different segments of customers.

---

Where is it used?

• Food product development (e.g., snacks, beverages)

• Tech products (e.g., what features to put in a new phone)

• Service design (e.g., hotel packages, subscription plans)

• Pricing strategies

The main types of conjoint analysis, from the traditional to the more modern ones are:

---

1. Traditional Conjoint Analysis (Full-Profile Conjoint)

• How it works: Respondents are shown complete product profiles (i.e., all features filled in) and asked to rank or rate them.

• Example: Rate this coffee: Medium roast, Ethiopian, $5.

• Good for: Simple studies with a few attributes (like 4–5 features max).

• Limitations: Gets overwhelming if there are too many features (cognitive overload).

---

2. Adaptive Conjoint Analysis (ACA)

• How it works: Starts by asking individual questions about preferences for single attributes. Then, the survey adapts based on earlier answers, showing more tailored questions.

• Example: First asks if you prefer dark or medium roast; then later asks you to make trade-offs involving those options.

• Good for: Studies with many attributes (up to 20+), personalizing the experience.

• Limitations: More complicated to design and analyze; newer methods have mostly replaced it.

---

3. Choice-Based Conjoint (CBC) — the most popular today

• How it works: Respondents are shown sets of product options and asked to choose their favorite — just like shopping.

• Example: Which coffee would you buy: A, B, or C?

• Good for: Realistic simulation of market choices; can handle larger studies well.

• Limitations: You don’t get individual “ratings” for each feature directly; you have to model it statistically.

Here’s a simple visual example of what a Choice-Based Conjoint (CBC) task might look like:

---

Imagine you’re doing a study for coffee products.

A CBC task would present you with a choice set like this:

Question: Which coffee would you choose?
(You can only pick one option.)

---

The respondent doesn’t separately rank “roast” or “origin” — they make a holistic choice just like they would at a grocery store shelf or an online store.

 The software records the choice.
 After many choice sets (maybe 8–15), the system analyzes patterns to infer:

• How much people care about roast level

• How much extra they’re willing to pay for Ethiopian beans

• How sensitive they are to price changes, etc.

---

Real-world parallels:

• Picking between different laptop configurations online

• Choosing a meal option in a food delivery app

• Selecting between mobile phone plans

---

4. Adaptive Choice-Based Conjoint (ACBC)

• How it works: A mix of ACA and CBC — first asks about must-haves or deal-breakers, then adapts the choice tasks based on early answers.

• Good for: Complex, high-stakes products (like choosing a new car or a complicated insurance plan).

• Limitations: Longer surveys, but more engaging and higher quality data.

---

5. Menu-Based Conjoint (MBC)

• How it works: Lets people build their own “bundle” — like configuring a meal, a car, or a subscription plan.

• Example: Choose coffee size, roast, add-ins (milk, syrup), etc., and see total price adjust.

• Good for: Customizable products and bundled offers.

• Limitations: More complex to set up and analyze.

---

Quick Table Summary:

---

Quick Note:

Today, Choice-Based Conjoint (CBC) is the most widely used — it best mimics how people really make decisions in stores or online.
ACBC and MBC are growing fast, especially for products that need some personalization.