Ba 28it - Conjoint Analysis

This assignment is to be completed as a team product with your assigned study team. The assignment is due at the start of the class session on April 28 (Cohort I) / April 29 (Cohort II).

The purpose of this exercise is to learn Conjoint Analysis by doing it.

Conjoint analysis is a tool that is commonly used in the product development process to help achieve a number of goals including:

• understand the importance that consumers place on specific product attributes (such as style, weight, color, or price)
• understand the weight or utility associated with “levels” of specific attributes (such as 3 pounds versus 5 pounds, or red versus blue)
• make predictions about preferences for combinations of attributes
• understand and predict consumers’ willingness to make trade-offs for features and attributes
• predict product demand at varying prices
• predict market share for specific product configurations  

In this exercise you will run develop, administer, and analyze a questionnaire with a conjoint design for a product or service of your choosing.  In class we will guide you through an example of this.  Consult the class handout for details that will help you with this assignment.

I.  Step One - Study Design.

1. Choose a product or service of interest to you.

2. Identify the three or four attributes that you, as managers, believe are most important to potential consumers of this product.

For example, if your product is a restaurant, your attributes might be cuisine, type of service, distance from campus, and price.  

As noted above, for the purposes of this exercise, you must narrow your study to what you see as the three or four most important attributes. In addition, you must include price as an attribute.  

3. Identify “levels” for each of your attributes.  

“Levels” are the values that each attribute can take on.  For example, the attribute of cuisine might include Mexican, Thai, or Bar-B-Que food.

For this exercise, your attribute of PRICE must have three levels (three different dollar values). All of your other attributes may have two, three, or four levels, as you see fit.  

4. Using one of the supplied experimental designs (see Appendix) create a questionnaire composed of an array of product profiles that can be rated on a 1 to 7 scale.  

NOTE: Be aware that designs provided in the appendix are specially generated “fractional factorial designs.” A properly created fractional factorial design reduces the total number of profiles that each participant must evaluate, while still allowing for successful analysis of main effects.  Other designs may not allow you to draw conclusions from your data.  

If none of the available experimental designs fit your set of attributes and levels, you may either adjust your attributes (more or fewer attributes / more or fewer levels) to fit one of the experimental designs or contact me to get an appropriate design.

Below is an example of a design for the restaurant example (see Appendix).  Note how 3 X 3 X 2 X 3 yields 54 possible combinations.  This fractional factorial design has reduced this to just 18 evaluations.  Every respondent will evaluate every one of the 18 combinations.

[pic 2]

II. Step 2: Data Collection.

Do your best to administer your questionnaire to individuals who are representative of your likely target consumer.

If this is not possible, you may use a convenience sample of people you know, for example, but you must be prepared to state this extreme limitation of your results.

Whichever route you choose, you must collect responses from a minimum of 10 consumers.

You will probably wish to collect this data online (using a tool such as Qualtrics or surveymonkey.com). However, you may use pencil and paper surveys if you choose.

III. Step 3: Data Analysis.

1. Using regression, estimate the part-worths for each of your ten respondents.

1. Choose one respondent and provide a copy of the regression output.
2. What are the part-worths for this respondent?
3. What are the best and worst possible products for this respondent?
4. What is the relative importance for each attribute for this respondent?

1. Plot the part-worths for the price variable for five respondents. Are these plots consistent with your expectation about the impact of the price in this product category?

1. Generate two new product profiles that did not appear on your questionnaire and call these profiles Product X and Product Y.

1. Assume that these are the only products in the market place.

Based on your ten respondents, what would you estimate the market share to be for each of these products? (If there is a tie for any respondent, then assume that individual would choose each product 50% of the time and assign half credit to each product for that respondent).

1. Your company is interested in introducing a single new product that will compete with the above products from your competitors; X and Y.
2. Generate three additional product profiles (A, B, and C) that are different from X and Y. Assume that these are three products that your firms’ R&D department considers feasible. Based on the ten respondents, answer the following:

1. One by one, determine what the market share would be if you introduce, A, or B, or C into the market where X and Y already exist. What would your market share be for each?
2. From part a, imagine that you did introduce your top product (A, B, or C) into the market with X and Y.

1. What happens to the market share of X and of Y?
2. What is the price elasticity of this product in the market with X and Y?

1. Now imagine that your team will introduce a second product into the market (one of the remaining two from A, B, or C)

1. Which would you introduce?  
2. What market share would it have?
3. How would this affect the market share for X, Y, and your own existing product?  

III. Step 4: Report.

1. Create a single document for your team. Only one assignment per team needs to be submitted. Be sure to put the names of all of the team members on the document.

2. Provide a description of the product you have studied, and its attributes.

3. Provide answers to all parts of all five questions from Step 3 (Data Analysis questions above).

4. Provide a paragraph in which you describe what you see as the most important information you learned from this research.  I don’t want to know why conjoint analysis is a valuable tool. I want to know what advice you would give to a manager if you had conducted this study as a consulting project.