A computerized reward mechanism system is proposed which makes use both of presently collected data relating to the customer, and a database of historical data relating to the customer. Using this information, the system performs a customer preference analysis, which results in selecting rewards which are adapted to the customers' personal needs. In one form, the system is positioned at a shopping center, and includes cameras for recognising past customers by image processing. This may be done without requiring customers to participate in a process of signing up to use the system. The presently collected data include responses to questions put to the customer relating to his present shopping experience. Customer honesty is considered in the step of determining which rewards to offer the consumer, to encourage the consumers to express their feelings more clearly.