Conducting a Conjoint Analysis: Definition & Examples

Conjoint analyses are usually carried out to determine consumer preferences. An online survey, for example, examines which individual features and which combinations of features matter to consumers when they assess a product as a whole. In product development, for instance, this research method measures which product characteristics are seen as particularly important or irrelevant and which combinations the target group prefers. This shows at an early stage of development which product variant consumers favour. Assessing customers’ needs and expectations during development, for example through customer surveys, is part of acceptance management.

Conjoint analysis: definition and explanation

Conjoint analysis is a multivariate method for measuring consumer preferences. It is used, for example, to measure how much individual product features matter from the consumer’s point of view. Alongside traditional conjoint analysis, which has been used in market research since the 1970s, several alternative conjoint methods now exist, such as adaptive and choice-based conjoint.

The term goes back to “conjoint measurement” (conjoint = joined): a product’s features are not rated one by one but together. In a conjoint analysis, respondents are therefore shown complete products with different characteristics and asked to choose the one they prefer.

Put simply, conjoint analysis lets potential customers choose between product alternatives, much as they would before making a purchase in a shop. They are shown a product whose attributes vary in their levels. This reveals consumer preferences and the utility of each attribute level.

The most widely used variant today is choice-based conjoint (CBC). It is often run with an online questionnaire in which product alternatives are presented in pairs or groups for respondents to evaluate. The variants can be shown as text, illustrations, photos and/or video. Respondents are asked to compare the alternatives and choose the one that suits them best.

The survey results are then analysed statistically and inform both the later stages of product development and the marketing strategy.

Conjoint analysis examples

Conjoint analyses are often carried out before a new product launch to determine the part-worth utilities of its components. If a food manufacturer wants to launch a new product, for example, a conjoint survey can reveal in advance which combinations of attributes potential customers prefer. For a new type of bread, the attributes could include the type of grain or gluten-free flour alternative (levels: e.g. spelt-walnut, buckwheat, almond-maize flour), the product name, the price or the pack size. In an online questionnaire, respondents would see one combination of these attributes alongside other product alternatives and pick their favourite. This shows which combination comes closest to what an individual respondent, or the majority of respondents, wants.

Conjoint analysis can also be used for pricing research (price-response function, price sensitivity etc.).

Another example: a manufacturer could use conjoint analysis to have prototypes of a new washing machine rated. The prototypes shown in the questionnaire would differ in the levels of individual attributes such as energy efficiency class (A, B, C), programme options (with/without water protection, with/without eco button etc.), load capacity and price. A customer survey would then show which combination of levels best matches respondents’ wishes and expectations. Based on the consumer preferences collected, the manufacturer could adapt the product to its target group or optimise its pricing.

Use our free conjoint analysis questionnaire template as a starting point. All questionnaire templates can be copied to your LamaPoll account with a single click and adapted as you like.

Conjoint analysis tool

An online tool makes it easy to run customer surveys for a conjoint analysis. The product alternatives are compared side by side in an online questionnaire so that prospects and potential customers can rate them. Survey invitations are sent to the relevant target groups by email or shared as a link on social media. Alternatively, you can embed the questionnaire on your own website.

Develop the ideal product based on consumer preferences collected in advance. Measure how much individual features matter and have combinations of them rated. Use customer surveys for conjoint analysis to make well-founded decisions!

How to conduct a conjoint analysis in 5 steps

1. Define attributes and levels: Choose a small number of attributes that really drive the purchase decision (e.g. price, size, material), each with two to four levels. Too many attributes overwhelm respondents.
2. Build choice sets: Combine the levels into realistic product variants and present two or three of them side by side in each choice set.
3. Set up the questionnaire: Explain the task in a short introduction, present the choice sets as questions with text or images, and add demographic questions so you can break down the results by target group.
4. Invite your target group: Send the questionnaire by email, share the link or embed the survey on your website.
5. Analyse the results: The results show how often each variant was chosen. To calculate part-worth utilities, i.e. how strongly each individual level influences the choice, export the data, for example as an SPSS or Excel file, and analyse it in statistical software.

Benefits of conjoint analysis

Conjoint analysis offers a number of benefits. Especially in its choice-based form, it can be used to forecast purchase probability and draw conclusions about consumer preferences.

Conjoint analysis allows you to:

• collect realistic data during product or service development
• determine the part-worth utility of individual levels
• draw conclusions about consumer preferences
• estimate demand
• forecast buying behaviour and purchase probability
• optimise products or services based on consumer preferences
• determine price elasticity
• tailor your marketing strategy to your target groups

Limitations of conjoint analysis:

• Only a limited number of attributes can be tested meaningfully.
• The purchase situation is hypothetical, so actual buying decisions may differ.
• Reliable results require a sufficiently large sample.
• Calculating part-worth utilities requires statistical expertise.

Create a conjoint questionnaire