Market Understanding:
Conjoint Techniques

About

As an addition (or alternative) to asking respondents to make a self-assessment about key attributes, we can ask them to make choices. This is the basis of conjoint. Making choices is something everybody does as part of their day-to-day lives; it is a natural thing to do. From this information we can infer the attributes which contribute to the choices they make. If asked well, it has the potential to be very accurate – avoiding, for example, some of cognitive pitfalls of self-assessment (respondents wanting to make themselves look good/professional; contrived consistency of responses; linear relationships which do not take account of natural thresholds and so on).

To get the best out of this method, the conjoint should be designed well – and the piloting process provides an invaluable opportunity to hone the questionnaire. There are five key factors that influence the correlation between stated intentions to act and actual behaviour.

  • The subject should be specific
  • The action should be specific
  • The context should be specific
  • The time should be specific and proximate
  • The intention should be personal

How we can deliver

Conjoint models are continually being improved such that they are:

  1. more natural for respondents to complete,
  2. tailored to the task required, and
  3. designed to provide more useful output.

The core output is to provide utilities of importance. In addition, the Latent Class conjoint models that we would propose to use can provide utilities on a respondent-by-respondent basis which can lead to a “needs-based segmentation”.

We will tailor the conjoint approach to the specific needs of a study. We have experience of choice-based conjoint, brand-price trade-off and MaxDiff. We work with a couple of specialist companies which gives us statistical support in running conjoint projects.

Part of Huntsworth Health | Privacy Statement | Copyright © 2010 Brand Health International