Asking questions and collecting answers—these are fundamental building blocks of any primary market research function. Computer-based interviewing has been our specialty for nearly 25 years. Our software may be used for collecting data over the web, or using computers not connected to the web (CAPI). Our interviewing tools support all the standard question types, including skip patterns, data piping, and dynamic list building.
What attributes are most important for my product? How can I design products and services to compete successfully in my market? How should I price my product? What about line extensions? These are key product design questions.
Over the last few decades, conjoint analysis has become the premier market research methodology for studying how buyers value the characteristics (attributes) of products/services and for predicting buyer behavior. Perhaps the most valuable aspect of conjoint analysis is the strategic what-if market simulator.
As researchers, we're constantly being asked to measure things, such as brand preference, the importance of product features, job-related factors and benefits, the impact of product packaging, etc. Sawtooth Software has developed a powerful system for scaling such items, called MaxDiff (Maximum Difference Scaling).
Many strategic business decisions are based on segmenting the market and determining how to reach target segments effectively. Good target segments are those that are substantial (large enough size), stable, and that reflect a good match in terms of buyer needs/desires for the firm's offering.
Pricing research is often recognized as one of the most difficult to execute well. Simply asking respondents how much they are willing to pay lacks realism and can lead to bargaining behavior. Using real sales data to develop price sensitivity curves can be problematic, as prices do not necessarily vary independently and often, and the many variables that affect buyer behavior (such as competitive effects) cannot be perfectly controlled.
Menu Choice and Bundling
Many products and services are offered using menus, where buyers can select from one to multiple options in configuring their ideal choice. Buyers are often offered fixed bundles of options as well as other options that can be offered a la carte.
A typical challenge we face as researchers is to estimate a variety of weights (such as utility scores, coefficients, or attribute importances) using a limited amount of data. A relatively new statistical methodology called hierarchical Bayes (HB) improves these estimates, leading to greater stability and validity.