Abstract: Nowadays, "touch the Internet" is becoming a key word for the Chinese brick-and-mortar retail industry, with dramatic changes of people's consumption lifestyle brought about by the Internet technology. Multi-channel retail model combined with traditional channel and Internet has become a standard mode for the retail industry. Under multi-channel retailing environment, consumer channel decision-making has changed fundamentally. The behavior that using different channels to buy becomes more and more popular. However, in the process of cross-channel buying, companies may also lose potential customers. For example, a consumer may enjoy the fitting service and personalized recommendations provided by a physical store, but end up purchasing clothes from the Internet, which seriously hurt the enthusiasm of traditional channel service personnel. Therefore, "how to guide consumers to migrate to the channel benefitting enterprises" becomes one of the important problems for multi-channel customer management. Firstly, a concept model of consumer channel selection mechanism including both search period and purchase period is established, based on the perceived value theory and trust transfer theory, specifically bringing products' search and risk characteristics into the model, channel-attributes, channel lock-in and cross-channel synergy's effects on consumer search intention and buying intention, and also, product search and risk characteristics' moderating functions to them are analyzed. Secondly, four simultaneous equations including consumer searching intention for traditional channel are listed, consumer searching intention for Internet channel, consumer buying intention for traditional channel and consumer buying intention for Internet channel are listed. Based on literature review, each variable of the equations is measured. Then data is collected by questionnaire survey. After the analysis of data's reliability and validity, parameters of consumer channel selection intention models are respectively estimated by 3SLS estimation using Stata12.0. Thirdly, consumer channel selection intentions about overall product categories and high-risk search product, low-risk search product, high-risk experience product, low-risk experience product are respectively analyzed. At last, conclusions and implications are put forward.
Some interesting results have been found. Firstly, channel-attributes, channel lock-in and cross-channel synergy affect consumer search intention and buying intention significantly. Secondly, for high-risk search products (such as mobile phones and personal computers), low-risk search products (such as books and stationery), high-risk experience products (such as clothing and cosmetics) and low-risk experience products (such as toys and snacks), there are obvious differences in channel search and buying attributes, traditional channel and Internet channel lock-in and cross-channel synergy's effects on consumer channel selection intention. Channel search attributes such as information availability, search convenience, assortment and social interactive affect consumer search intention positively, but search effort affects it negatively. Channel buying attributes such as service quality, purchase convenience and enjoyment affect consumer search intention positively, but purchase effort, purchase risk affects it negatively. Traditional channel has lock-in, but Internet channel hasn't. Otherwise for low-risk search products, Internet channel also has lock-in. For high-risk products, traditional channel and Internet channel has cross-channel synergy. For high-risk search products, there is some positive cross-channel synergy from Internet search to store purchase. For high-risk experience products, there is some positive cross-channel synergy both from Internet search to store purchase and from store search to Internet purchase. For low-risk experience products, there is no significant cross-channel synergy between the Internet and the store.
All of these throw lights on practical implications for enterprises' channel construction, increasing customer stickiness and managing multi-channel customer effectively.

Key words: multi-channel retailing, consumer searching intention, consumer buying intention, channel attributes, product categories characteristics