Recommendations

What might be called "recommendations" from the shopper's perspective is what is broadly known as "marketing" from the shopkeeper's perspective. Shoppers often come to a retail Web site (or a conventional store for that matter) with a particular product in mind, and sometimes even the brand and model. Marketing is the process of getting the shopper interested in products that they did not necessarily already intend to purchase.

Promoting accessories along with their associated products is a form of marketing that is sometimes called cross selling. Getting a shopper to change their mind and buy a model that has more features is a form of marketing that is sometimes called up selling. Offering discounts of various sorts is a form of marketing, and suggests to the shopper that now is a good time to make the purchase.

Advertisements, which we all know are quite common on Web sites, are a less directed form of making recommendations to shoppers, but they certainly qualify as marketing. When information about the shopper is known, such as their age or gender, that knowledge can be used to display advertisements to which they are more likely to respond. This is sometimes called targeted advertising.

Direct mail, often thought of as "junk" mail from the shopper's perspective, must be a statistically effective way to market products, or there wouldn't be so much of it. Most of us are already familiar with the e-mail equivalent of direct mail. This is certainly a marketing technique that is often employed in conjunction with a retail Web site. Notably, the e-mail version of direct mail provides a more convenient way to request removal from the mailing list, usually a link in the e-mail message that you can click; this particular functionality is often called "opt out."

Some popular Web sites use predictive algorithms so that they can make recommendations along the lines of "other people who bought this product were also interested in the following other products." These recommendations are based on relationships that may not have been natural to foresee, but which are based on the actual buying patterns of previous customers. These algorithms are complex and proprietary, and show how the tracking of Web site activity can be exploited to increase future profits.

Commerce Server provides a variety of functionality to support all of these various methods of marketing. They can be used in various customized combinations as additional ways in which to gain an advantage over competitors. These methods are summarized in the following table:

Marketing Method	Description
Cross sells	Recommending related products, such as accessories
Up sells	Recommending upgraded versions of a product
Promotions	Prompting immediate purchase based on discounted or otherwise attractive current prices
Advertisements	Recommending unrelated though potentially targeted products, often receiving revenue in return
Direct Mail	Send potentially targeted e-mail inviting purchases at the Web site
Predictions	Recommending other products based on other customers buying habits

The next three sub-sections discuss the support provided by Commerce Server for recommendations to shoppers, also known as marketing, in each of the three types of functionality: user interfaces, Web sites, and objects.

Previous Next