Much market research is aimed at finding out how consumers perceive goods and services and what they want. This is essential in designing marketing campaigns, to understand what approach to the consumer might work the best.

However, the earlier stages of product development and pricing are often what makes or breaks a product. It is no good having a product that is too expensive because costs are going into features the consumer is unwilling to pay for. It is equally bad to miss opportunities where the consumer is willing to pay significantly extra for easy to provide features.

At the centre of these issues is a simple question. If presented with a range of products at a set of prices, which, if any, will the consumer purchase? Questions like this occur every day. What brand of car to buy? Whether to travel by car or train. What party to vote for? Whether to pay more for environmentally friendly products or locally made products. Which pair of shoes to buy?

Data Analysis Australia believes that where the consumer faces a choice, the research design should replicate that choice process. Forcing a survey respondent to make a thoughtful choice is critical. In its simplest form a product is presented to a sample of consumers with a range of prices - each consumer seeing just one price - and they are asked whether they would purchase it.

As a rough guide, if more than half the sample is willing to pay a certain amount then it suggests the price can be increased. A more sophisticated analysis can tell much more, such as determining how sensitive consumers are to prices, so that the optimal price can be chosen.

Central to this is the emulation of the actual decision process. When the real decisions are more complex - such as between multiple options - the research questions must be more carefully crafted to be as realistic as possible. Consumers in the sample must be faced with realistic choices and realistic information. This is a challenge in the design of the questionnaire or interview.

The overall design of a study usually considers a large range of options, too many for all options to be presented to any one person. Effective design presents carefully chosen "choice sets" to each person. Done well, every question yields rich information on a number of factors. Done poorly, it can be inefficient or, worse still, give ambiguous results that are close to useless.

The secret to success comes from pure mathematical structures called, not surprisingly, finite designs and are based on concepts developed by French mathematician Galois in the early 1800's. The reason why Data Analysis Australia is able to design such efficient surveys is our combined strength in mathematics and the understanding of market research objectives.

The analysis requires the right statistical models. These models must reflect the structure of the data and the decision process. For example, sometimes it is appropriate to analyse in terms of the sequence of small decisions that then lead to the real drivers behind the reasons people make their choices.

These techniques go under many names, often associated with a single area of application. Economists talk of contingent value analysis. Market researchers talk of conjoint analysis. Transport planners talk of modal choice analysis. Still other areas talk of trade-off analysis. The names do not matter. Even though practices in these areas have some differences, the similarities are greater since all require the intelligent and creative use of the same mathematical and statistical tools.

For further information, please contact Data Analysis Australia at **daa(at)daa.com.au** or phone **08 9468 2533**.

*October 2009*