Models
|
Conjoint Analysis Introduction Products usually have many attributes, such as price, color, style and its unique function, etc. Then which attribute of products are the one that consumers concern most? Further analyzed, the essential of this question is: How important is each attribute to consumers? What specialties that the product has will be most possibly accepted by the consumers with the same cost? For such kind of questions, qualitative research method as well as the method of getting direct statements from respondents were usually adopted in the traditional market research. These methods, however, could not get a precise analysis results. Conjoint Analysis is produced aiming at solving this problem. It can simulate the trade-off process that the consumers give up some attributes for satisfying some demands in reality. Its results would be more objective and accurate.
CBR's Conjoint analysis CBR adopted the most authoritative Conjoint analysis software-Sawtooth- package to do multiple types of research analysis including Adaptive Conjoint Analysis (ACA), Choice Based Conjoint (CBC) ,etc. · CBR Conjoint analysis can be implemented by PAPI or on computer interface. Sample for D Segment Car Buyer CBC Utility Estimation Through conjoint analysis, we can accurately know consumers' preference to each type of product, each price level, and each brand, and which attributes play more important roles, etc.
Market Share Simulation
Adopt different research model by different research term? CBR provides the clients different conjoint analysis models according to their specific research objectives. ： · CBC is a research model based on the consumers ' choice ， by which the conner only needs to simply make a choice in a series of virtual products to analyze consumers' internal preference . · CBC is suitable for many research circumstances. Advantages: The testing scene is much more real and closed to consumers' actual choices. Disadvantages: It usually needs more samples to get accurate results for assuring the precision of research. · ACA is a model very suitable for multi-attributes (more than 6 attributes), or attribute levels, which can save the time and relieve burden of respondents. Also, it can get satisfying results even through few samples. · Since ACA has a certain of warp to estimate price factors, it is not usually for special price test.
GAP Model Introduction Gap model is used for analyzing consumers' demands by CBR. Its essential idea is to seek the key demand points that drive consumers to purchase the products by analyzing the relationships between the present demand gap and future purchase, and forecast possible purchasing rate under different gap level according to demand gap size.
Case -What factors lead consumers upgrade present DC Research indicates: The main reason why consumer upgrade the present DC products is that they are dissatisfied with pixels, and pursue zoom capacity.
The gap means the ideal DSC's configures of the respondents minus their current DSC ‘s configures.
Decision Tree Model Characteristics of decision tree analysis · Decision Tree Analysis is one of basic modern data mining methods, which synthetically compare independent variables and automatically choose those variables that affect objective mostly, and therefore find the optimal classifying mode. CBR successfully developed a series of application modes based on decision tree technology, which can help the clients to do the followings. · Market segmentation optimization ： Help clients choose the optimal market segmentation based on decision tree results. ·Gain Analyses ： Analyzing the potential and sales efficiency of each segmented market to help the clients' choose the optimal segmented market. · Long term application ： Building a defining regulation for market segmentation based on decision tree to define the classification of unknown objectives and therefore strongly supporting the clients' data base marketing. Illustration of Decision Tree Analysis Phase 1. Choose all possible segmented variables à choose all target variables which support decision Phase 2. Choose decision analysis methods à system operation, automatic inspection Phase 3. Export decision tree graph à Export Gain Table
Research Case · Clients will face various possible segmentation modes in some IT product research: geographical segmentation, income segmentation, gender segmentation …. Which kind of segmentation is most beneficial for finding potential consumers? · To find the answer, we choose whether to purchase in next 6 months or not as target variable, and take purchasers as the research target. · For independent variables, we add as many demographical variables, which possibly affect purchasing behavior, as possible: age, income, city, gender, marriage, education … · Please find analyzing results vide post.
[back] |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
