Althuizen - Analogical Reasoning As A Decision Support Principle For Weakly- Structured Marketing Problems

ISBN
   

Soort boek
Onbekende bindwijze

Many marketing problems are weakly-structured. From the psychological literature, we know that analogical reasoning is an effective problem-solving method in weakly-structured decision situations. That is, when confronted with such a problem, one of the first things managers will naturally do is to search their memory for previous, similar experiences (or cases) that could help to solve the problem at hand. Case-Based Reasoning (CBR) is a state-of-the-art artificial intelligence technique that mimics this kind of human reasoning and can be used to put the vast amount of experience-based marketing knowledge into action for decision making. In a series of experiments, we investigate the effectiveness and efficiency of CBR system availability under different conditions. We use the design of sales promotion campaigns as our weakly-structured application domain in marketing. This dissertation shows that analogical reasoning does help to improve marketing decision making. More specifically, we demonstrate that a CBR system helps to design better, more creative sales promotion campaigns and helps to design them more efficiently. Furthermore, we show that the effects of CBR system availability are dependent on: (1) the content of the case-base (i.e., near analogies versus far analogies), (2) the size of the case-base (i.e., large size versus standard size), and (3) the characteristics of the decision maker (i.e., creative ability). Interestingly, the effect of CBR system availability on solution quality is largest for decision makers with a low creative ability (i.e., a compensation effect). Moreover, we show that decision makers have difficulties in recognizing the contribution of the CBR system to the objective quality of their solution. Based on these findings, we provide a number of recommendations for managers, such as “capture, store and make knowledge available in the form of cases” and “different types of decision makers need different types of CBR systems”.