Blockbusters and the long tail (part 5)

However, most studies of the sort described previously have been conducted in artificial settings where consumers have very little outside information about the products they are evaluating. For that reason, we decided to investigate whether herding would persist in real-world markets in which consumers could easily collect outside information about the products they were evaluating. To do this, we partnered with a major cable company to conduct an experiment using their sales platform. We added a new menu to the company’s video-on-demand service that displayed the most popular movies according to other consumers’ recent ratings.

In the default case, this menu displayed fifteen movies in descending order of the number of likes each movie had received from earlier users. However, at discrete times during the experiment we reversed the placements of two movies on the list. If user behaviour is strongly influenced by the opinions of the herd, we would expect that users would rely on this incorrect information about a movie’s “likes,” and that the artificially promoted movies would remain in the artificially higher position or could even increase in popularity as they gained followers and exposure.

We ran the experiment for six months in 2012, during which time more than 22,000 users purchased movies from our experimental menu of options. Our results showed little evidence of long-term herding behaviour. When a movie was reported to have more or fewer “likes” than it really had, subsequent reviews by users caused it to return to its original position quickly. Moreover, better-known movies returned to their original positions more rapidly than lesser-known movies. In short, our experiment showed that consumers were less likely to follow the herd when they have access to outside information about the products they were evaluating – as might be expected among online consumers who can easily gather information about millions of different products.

Increased product variety, improved search tools, recommendation engines, peer reviews, and increased product information each play a part in driving online consumers toward niche products. But there is one other factor to consider: whether the anonymity offered by online transactions might change consumers’ behaviour by reducing their inhibitions.

Avi Goldfarb, Ryan McDevitt, Sampsa Samila, Brian Silverman analysed this effect in two contexts: purchasing alcohol and ordering pizzas. (We know that alcohol and pizza aren’t entertainment goods, but stick with us here.) The researchers found that when consumers purchased alcohol using a computer interface, they were more likely to choose products whose names were hard to pronounce than they were when ordering over a counter from a human clerk.

Similarly, when ordering pizzas through a computer interface, consumers were more likely to order high-calorie products and more complicated toppings than they were when ordering by telephone. Goldfarb et al. argue that the increase in online ordering of difficult-to-pronounce products probably occurs because of consumers’ fears of “being misunderstood or appearing unsophisticated,” and that inhibitions surrounding face-to-face orders for complicated, high-calorie pizzas probably are driven by consumers’ concerns about “negative social judgment of their eating habits” and “negative social judgment associated with being difficult or unconventional.”  

What do alcohol and pizza have to do with consuming entertainment products? Beyond the obvious demand complementarities, it’s easy to see how reduced social inhibitions in online transactions might also affect consumers’ choices for entertainment. Katherine Rosman made this point in a 2012 Wall Street Journal article titled “Books Women Read When No One Can See the Cover,” which documents the recent growth in demand for certain publishing genres. “Erotica used to be difficult to find,” she wrote. “Chains and independent bookstores might have carried a few titles, but they were hidden away, and inventory was scarce.” The anonymity offered by Kindles and other e-readers has changed all that. Think about the success of a long-tail product such as E.L. James’ 50 Shades of Grey.

Now, we understand why you may have choked on that last sentence. After all, 50 Shades of Grey isn’t a long-tail product. It has been translated into more than fifty languages, has sold more than 100 million copies, and has spawned a movie franchise. It’s a classic blockbuster! You’re right, of course – but you’re also wrong. In many ways, 50 Shades of Grey is a classic long-tail title. It was rejected by traditional publishing houses, it was brought to market not through a reputable print publisher but as a self-published e-book, and we wouldn’t be talking about it today if it hadn’t been aggressively promoted by passionate fans in online communities.

Here’s the rub: 50 Shades of Grey, like many other products today, has elements of both a long tail and blockbuster. And in spanning those two categories, it highlights the limitations of focusing on products rather than processes, at least when it comes to understanding how technology is changing the entertainment industries.