One of the featured speakers at Friday’s BISG meeting was Anita Elberse, associate professor at Harvard Business School, who challenged some conclusions of Chris Anderson’s long tail theory. Elberse said that while the growth of online retailing has resulted in the expansion of products that are available for sale in the long tail, there is little evidence to show that sales of niche products have significantly increased. “The tail is getting longer, but it isn’t getting fatter,” Elberse asserted, referring to Anderson’s contention that with more items in the tail, sales will increase.

Her studies of music and DVD sales found that hits still account for the vast majority of sales and that sales of niche products have fallen in some cases, and there are lots of products that simply do not sell. Contrary to assumptions that consumers who are “light” users of a particular form of media will be attracted by some niche offering, Elberse said these consumers most often buy a popular item. Someone who only reads a few books a year is most likely going to buy a bestseller rather than an obscure author few people have heard of. Lack of consumer awareness about different products is one reason products are in the tail, as is quality. “Some products belong in the tail not the head,” Elberse said.

Her advice for publishers is to not radically alter the resources they devote to developing products for the tail and to continue to concentrate on publishing books aimed at a wide audience. While the Internet has changed how people buy products, it hasn’t changed the laws of consumer behavior in which most people want to read or watch what is popular. She acknowledged, however, the retailers, particularly online retailers, need to carry niche products to help separate themselves from their competitors. But she noted that the niche items are likely to be bought by “heavy users” who are looking for new books (or other products) since they have likely bought many of the most popular items.

Jim Milliot
Publisher’s Weekly

Advertisement