Hi, I apologize that this question maybe half product promotion, but I really want to know what people think about the approach. I have got some positive feedback and potential sales cases for the module within track or ringtone online sales, but would very much appreciate opinion of this community. The idea is that we analyze list of purchase transactions (what customers buy as a single purchase: track1, track23, track76, ..., track17) and search for associations within purchases finding similarities and patterns. Thus we have two results: 1. we can say f.e. that if somebody buys trac23, track67 and track 44 then he will also buy track 17, and furthermore one can predict that if he buys track23, track67, track44 then he would also like to listen to track86. All of this comes from analyzing purchase transactions made by other customers. This can be done once a day f.e. or even live for a single customer while he makes his purchases on the site. This leads to the following scenario: an online customer buys one track, the script checks what other tracks usually come with this one and displays links to these tracks f.e. on the right panel attracking customer's attention. When the customer buys second track (not necesserely from the right panel) the procedure repeats and so forthe till he stops buying or exceeds "the knowledge" hidden in the purchase transaction data. This results to customers easier finding tracks of their interest and increasing purchases (they get links to tracks they are interested in, because the analysis is based on the purchases of the customers with the same taste). This module is implemented both as an online service or as an integrated module returning analysis results in XML format. What do you think about such module and its perspectives for online shopping? Thanks a lot in advance! P.S. Direct link to online demo and module download: http://www.sevana.fi/context_data_analysis.php __________________ Sevana Oy IT Solutions & Services http://www.sevana.fi
