Streaming Choices: Do We Follow the Big Crowd or Our Friends?

The ubiquitous community-wide customer rating is the biggest pull in helping us pick a movie or show

Deciding what to watch is becoming a bit of an overwhelming task, given the thousands of movies and television shows available on streaming sites such as Netflix, Hulu and Amazon Prime. You could spend an entire night just scrolling through all the options.

Cognizant that too much choice can quickly turn frustrating, streaming companies serve up user-based ratings and reviews. The social feedback we can lean on to help decide what to watch comes in two basic flavors: the “average customer rating” is a form of word-of-mouth feedback; data tracking “how many times a show was watched” fall under the “observational” cue whereby we might base our decision not on opinion, but on herd mentality.

We can take our word-of-mouth or observational cues from the entire community at large or from the site friends that some services encourage us to make through chat rooms and other features.

Given the high stakes of customer satisfaction in the competitive streaming world, understanding what specific type of nudging might be more (or less) effective could have a tangible payoff for streaming services.

Mina Ameri and Ying Xie of the University of Texas at Dallas and UCLA Anderson’s Elisabeth Honka make a case for streaming businesses to continue to focus on customer ratings from the broad community. The researchers found that, more so than the feedback from a personal network of cyber-friends, the community-wide word-of-mouth opinion was the most effective factor in increasing viewership.

Ameri, Honka and Xie studied the viewing habits of users of the world’s largest Japanese anime (cartoon) website who were actively using the site’s social networking. The authors created a random sample of 5,000 users from the universe of nearly 380,000 socially active users, and studied their weekly viewing habits for 103 anime series, released between July 2012 and January 2014, that had at least 50,000 views. All told, there were nearly 22 million weekly viewing data points to work with.

The authors computed averages for the number of views, the average rating, and how many ratings each anime had, across both the larger community and within a user’s personal network. All three elements create a positive nudge across the community. Within a friends’ network, the number of views and number of ratings had a positive effect. Overall, community feedback is more effective than input from a personal network. The bigger the herd to follow, the better it seems.

They then predicted how a 1.0 percent increase in a variable — looking at both the community and personal friends’ networks — might impact viewing habits.

The average community rating has the biggest pull with viewers. A 1.0 percent increase in a community rating boosted viewing of a given anime by 0.6 percent. A 1.0 percent increase in the number of viewers triggers a 0.49 percent increase in viewings in the model. And a 1.0 percent increase in the number of community ratings produces a 0.03 percent increase in views.

There is a different ranking of the most effective factors within the personal network. Whereas a change in the average rating was the largest driver of viewer increases among the site’s entire community, it was the weakest factor when the authors applied the same 1.0 percent test to the personal networks.

A 1.0 percent increase in the average rating among the personal network of site friends produces a 0.01 percent increase in views. A 1.0 percent increase in the number of ratings within a friends’ network triggers a 0.27 percent increase in views, and a 1.0 percent change in the number of views results in a 0.21 percent pickup in views.