Conjoint analysis

Conjoint analysis – Definition

“What is a conjoint analysis?”

Conjoint analysis, also referred to as conjoint measurement, is a multivariate method used to determine customer benefit and willingness to pay.

In order to determine the utility of product characteristics for the customer and their willingness to pay, customers must rank various products with different characteristics. These different characteristics may, for example, be technical specifications, trademarks or additional services. The individual willingness to pay can be determined by considering the different customer preferences for the separate components. The advantage is that this selection procedure comes very close to a real purchase decision. The conjoint analysis offers a good basis for setting utility-based prices.

Conjoint analysis – fields of use

“When to use a conjoint analysis?”

Conjoint analyses are suitable for dealing with numerous questions in marketing. The specific fields of application of conjoint analysis in marketing are diverse and range from product development to pricing and market segmentation. The following questions, for example, can be explored with a conjoint analysis:

Field of use: Product development

  • Which product features do customers see as fundamentally important?
    (Example smartphone manufacturer: How important is the camera quality in the smartphone to customers?)
  • Which product features are more important than others?
    (Example smartphone manufacturer: Is camera quality or battery life in the smartphone more important to customers?)
  • How can differentiated package offers be optimally designed?
    (Example smartphone manufacturer: Do customers prefer a smartphone with a poor camera and long battery life or one with a good camera and short battery life?)

Field of use: Pricing

  • How can the price of an existing product or service be adjusted to generate higher sales?
    (Example: How does the sales volume of my product change if I increase or decrease the price by x%?)
  • How high is the willingness to pay for a better specification of a product attribute?
    (Example: An industrial goods manufacturer has made an existing product more powerful and now wonders what the additional willingness to pay of customers is)
  • How can a new product or service be priced in an existing market?
    (Example: a machine manufacturer has developed an additional service package and is now asking at what price the service should be offered)
  • How should a disruptive technology be priced?
    (Example: What is the willingness to pay for a novel disruptive product based on its relative performance characteristics?)

Field of use: Market segmentation

  • Which market share can a new product achieve?
  • Can clear market segments be formed on the basis of demand?
  • What is the optimal product or service for different market segments?

Conjoint analysis – Procedure

“How to conduct a conjoint analysis?”

For a conjoint analysis to be successful, the target image must first be defined as concretely as possible. Will the analysis be used for (new) product development or for pricing? Is the aim to increase product profit or market coverage?

After defining the target picture, the target group for the survey must be defined. It is important to ensure that the selected group of people / customers is representative and suitable for answering the initial question or for achieving the target picture.

The next step is to define the exact type of conjoint analysis. Depending on the objective, one of the following three methods is usually selected:

  • Classical Conjoint Analysis
  • Choice-based conjoint analysis
  • Adaptive conjoint analysis

Based on this, the relevant features and the respective characteristics must be defined. When making the selection, it is important to ensure that the characteristics and values are sufficiently meaningful and clearly distinguishable from one another. In addition, the number of features and characteristics should be based on the objective and the target group. As a guideline, a maximum of 5 – 7 characteristics with 2 – 4 values each should be defined. Each additional characteristic or expression increases the requirements for the size of the sample and the duration of the survey.

Subsequently, the selected characteristics and their proficiencies are combined into fictitious product variants. During the creation, non-logical variants should be excluded to ensure the feasibility of the product.

Depending on the target image, competitor products can also be included for selection in addition to fictitious product variants. Such a procedure extends the selection of product variants, but enables a more realistic decision scenario.

As part of an online survey, participants are asked to evaluate the different product variants along their own preferences. Here, attention must be paid to the exact question. Depending on the objective, the question must be carefully considered. For example, the questions “Which product variant do you prefer?” or “Which product variant would you be most likely to buy?” can lead to different results and may not produce the hoped-for results.

Finally, the analysis calculates the relative importance of the characteristics and assigns a utility value to each characteristic value. This reveals the partial benefit of the respective characteristic within the characteristic category. In the analysis, the utility values of the individual values of the characteristics are evaluated and assessed separately. The allocation of the individual utility values is based on multivariate statistics (likelihood method and hierarchical Bayes estimation algorithm). In addition, price-sales functions can be created, which are used to derive theoretical willingness-to-pay values.

The mathematical results of the conjoint analysis can now be used to answer the questions posed at the beginning and to derive appropriate measures for the portfolio, price or sales strategy. Depending on the objectives, optimal product variants can be defined, price adjustments determined or target segments identified.

Conjoint analysis – Example methods

“What types of conjoint analyses do exist?”

Classical conjoint analysis: trade-off method

  • Trade-off: for the evaluation of a stimulus, two product characteristics are evaluated using numerical values given according to preference order.

Classical conjoint analysis: full-profile method

  • Full-Profile: Subject receives two cards with full product (attributes) (which are also described); stimuli are then ranked and assigned a utility value (high = important/positive); partial utility values together then reassigned utility value.

Choice-based conjoint analysis

  • Selection of a product concept from a set of concepts; no evaluation of individual stimuli.

Adaptive conjoint analysis

  • Computers analyze the subject’s answers while the survey is still in progress, so that subsequent questions are adjusted based on previous responses.

Maximum-difference conjoint analysis

  • Test persons evaluate product or service by indicating which is perceived as the best and which as the worst product; especially used when product features contain only one characteristic (present or absent).

Limit conjoint analysis

  • Similar to the full profile method, with the addition that the test person places a limit card describing up to which rank a product would be purchased (should there be no product with higher preference).

Conjoint analysis – success factors

“What must be taken into account to conduct a successful conjoint analysis?”
  • Definition of a clear target image with a few but concrete questions
  • Selection of a representative target group and determination of the appropriate type of conjoint analysis
  • Definition of the relevant product features (max. 5-7) and characteristics (max. 2-4)
  • Modeling of a realistic decision situation
  • Design of a CI-conform survey with a professional look & feel
  • Emphasis on the protection of personal data and anonymized evaluation

Conjoint analysis: Care to find out more?

If you would like to learn more about conjoint analysis or have identified a specific application need for your company, please feel free to contact our experts!

Porträtbild unseres Partners Markus Pfeifer
Partner

Markus Pfeifer

markus.pfeifer@homburg-partner.com
Tel.: +49 621 1582-322

Potrtät von Ansgar Niehage
Consultant

Ansgar Niehage

ansgar.niehage@homburg-partner.com
Tel.: +49 211 598929-56

Enough theory? Interested in practical insights? Learn more about our industry-specific pricing expertise: