|
Conjoint analysis (Trade-Off)
Understanding and modelling the consumer decision process
When
choosing a product out of several, consumers opt for "the
best compromise". They choose the one that, based on
their perceptions, best meets their expectations. They therefore
give preference to certain product attributes to the detriment
of others which are less important for them.
The Trade-Off model, or conjoint analysis, enables us to measure
not only buying behaviour, but also to explain how and why
purchasing decisions are made.
Knowledge of the decision process obviously enables us to
influence future purchasing decisions.
Results from a conjoint
analysis
The product or service that is the object of the study
is broken down into characteristics or variables, each
with different modalities (e.g. the price variable might
have 4 modalities : 10, 15, 20 and 25 Euros).
Conjoint analysis enables us to determine:
• The relative importance of each variable in the decision
process.
• The appeal (utility) of each of the modalities
of the variables
In particular, it enables us to simulate the share
of choice of all products made up of a combination
of the variables, in absolute terms or within a competitive
environment. For example, for a Trade-Off on mobile telephones
with three variables - brand, number of hours in package
and price - we could simulate the share of choice for
each of the packages proposed by the different brands
and then evaluate all impact on share of choice of any
modification to pricing levels.
Choice of a conjoint analysis model
Several
conjoint analysis models have been developed. At Repères,
we prefer the CBC (Conjoint Based Choice) model which
builds in all the latest developments in Bayesian statistics
and means:
• A high degree of flexibility in the choice of variables
to be included in the analysis (up to 10 variables and 15
modalities per variable)
• The possibility to validate beforehand the statistical
robustness of the experimental designs and the expected
reliability of the results
• Particularly simple and realistic data collection:
Respondents simply have to make a series of ten or so choices.
For each decision, they simply have to say which product
they prefer out a reduced number (often 4 or 5) of propositions
• The possibility to use CAPI or online interviews
thanks to the simplicity of the data collection process
• Calculation of the utilities at an individual level
(i.e. for each respondent), enabling us to identify typologies
of respondents and thus uncover homogenous groups of consumers
in terms of expectations
Applications
•
Identification du positionnement sensoriel le plus attractif
et mise au point du produit correspondant
• Création d’une gamme couvrant la diversité
des attentes en élaborant le produit optimal pour
chacun des segments de consommateurs ayant des préférences
homogènes
• Suivi qualité de la production : projection
sur la carte sensorielle de différents lots de production
d’un produit et vérification de l’adéquation
à la cible sensorielle recherchée
Fields
of application
We
use the Trade-Off method to:
• Optimise a marketing mix so as to meet the
expectations of the target population better
• Construct a range of products to cover a diversity
of expectations
• Predict the potential in terms of share of choice
of a new product in an existing market and identify the
means to optimise it
• Evaluate the price elasticity of a product or service
and determine the optimal price strategy
• Etc…
The Trade-Off method requires interviews no longer than
20 minutes and so can be built into bigger studies (U&A,
image study) for a limited extra cost.
Repères
and the conjoint analysis
•
We have been using the Trade-Off method for over 10 years,
for many industrials in both the product and service sectors
• Our expertise means that our clients are guaranteed
optimal and always appropriate usage of the method
(in particular cases we may not recommend this approach
if it is not the right one)
• Complementary to the results from the study, we also
deliver user-friendly software to our clients that
enables them to simulate all possible scenarios (variable
combinations)
Contact
reperes@reperes.net
|
