In its second-largest acquisition, Amazon buys the company for $970 million.
(Page 3 of 3)
The same retailer wanted to know if a cross-sell on its site was effective. The tool showed that although some shoppers followed the link from the main product to the cross-sell, the numbers weren’t very high. Since the cross-sell display was not located near the main product display, the data raised the question of whether a closer link between the main product and the cross-sell would make the cross sell more effective.
Web site analytics can slice and dice customers’ clickstream behavior in enormous detail, answering a near-infinite number of questions that could theoretically drive the same number of online merchandising decisions. If an item is clicked often but carted infrequently, for example, it may mean it’s too expensive. Should the merchant drop the price or move it to less valuable real estate? Is the click-to-buy ratio for an item high when shoppers find it by site search, but low when they find the same item in summer specials? That could suggest the item’s strongest market is really among those already searching for it, and that other items are better suited to the impulse-buy mentality of special promotions.
With all the possibilities, merchants with new access to information on shoppers’ on-site response could re-merchandise on-site every day or even every hour. But as a regular occurrence, experts say, that’s counterproductive-not to mention a drain on staff resources. “The danger of too much data is wasting time on stuff that’s not a big deal,” Heller says.
A few questions
To help retail clients prioritize merchandising decisions and use data selectively, Heller advises them to start by making changes at the segment level rather than at the individual level. One client, for example, came up with a list of several dozen business questions for DoubleClick’s analytics tool. To narrow the field, DoubleClick ranked the questions by significance of impact. An initial analysis showed that one question, for instance, involved a situation that would have an impact on about 10% of the site’s revenues and 14% of its traffic. “That was a big one, so we put it at the top of the list,” says Heller. “We found that another question would affect about 1% of its revenue and about 0.5% of its traffic, so that went to the bottom of the list.”
Heller advises clients looking to prioritize the use of analytics to improve customer experience on sites to look at particular sections of the site to determine whether they are generating the desired revenue, then decide if the section could generate more revenue with a different treatment. They can next look at different audience segments, seeing if shoppers are following the path toward completion of an intended act, such as making a purchase, and consider alterations there. Another key way for retailers to use analytics is to identify the top most-browsed products that aren’t being purchased, which may suggest that pricing changes or promotions are needed to draw customers more effectively.
Retailers have always known that effective merchandising is merchandising that speaks immediately to the need of customers. The challenge isn’t so different online, but web speed gives a new definition to the term “immediate,” and it requires that merchants learn a new language and master a new tool set to be effective.
“Retailers have been doing merchandising for hundreds of years in stores, and they do a really great job of tracking their customers around a store, setting up a plan-o-gram to make sure products are where they will move most effectively, and populating end-decks with the right items,” says Todaro. “The big thing for them now is to be able to take that same concept that they use in-store or the way they’ve laid out products in a catalog, and apply that same capability online.”