The e-retailer reports a $126 million net loss, stemming from a $640 million year-over-year increase in spending in the quarter on technology and content ...
Using dynamic merchandise presentation to cross-sell has limited effect when the goal is add-on sales versus core purchases, Jupiter finds. A more productive approach: dynamically generate content for e-mails pushed out to customers by segment.
A full 76% of online retailers use at least one method of cross-selling to suggest products to online shoppers – but only 9% of online buyers use suggestions in researching and buying the product online, according to new findings from a Jupiter Research report, “Dynamic Merchandising: Targeting Merchandising Through Segmentation.”
The personalized shopping experience potentially afforded by dynamic merchandising faces challenges to effective execution on a couple of fronts, according to the report’s author, Jupiter analyst Patti Freeman Evans. One is the nature of the product suggestions themselves. Most retailers polled reported using them to push add-on sales, not the core purchases that bought consumers into the store to begin with. “This method will have limited effect because only 25% of online buyers said they buy products they didn’t intend to buy, even when presented with an attractive offer,’ says Freeman Evans.
Freeman Evans adds that ROI on fully automated dynamic merchandising, which infers from customer behavior which merchandise or content is most suited to a particular customer, is as yet so uncertain that Jupiter does not recommend it as a means of site personalization. However, with better analytic tools available for customer segmentation, triggered e-mail campaigns represent a good way for retailers to initially test effectiveness of dynamic merchandising. Research studies suggest that the use of web site clickstream data as a targeting tool in a dynamically generated e-mail increases relevance and produces 16 times more improvement in net profit than that generated by broadcast campaigns, Jupiter found.