The more you have choices, the more you struggle to choose. That is the paradox of choice which creates decision fatigue for Netflix users. To fight against this phenomenon, Netflix announced a new shuffle feature to facilitate the decision: “Play Something” to launch automatically something to watch. Recommendations are still made on series and films you have already watch on Netflix: if you watched them, you liked them, so we recommend you something similar. But is “watching” the best qualitative indicator? How Netflix could recommend content differently to avoid decision fatigue?

Why do you take so long to choose something to watch?

If you are a Netflix user, you have already experienced (at least once) the famous “they are too many choices” while you are navigating on the platform and keep scrolling. So, you have experienced the Hick’s Law on Netflix, and certainly in other moments of decision making, but especially online because you have a very wide range of products. According to the Hick’s Law : the more options you have, the more you take time to “respond” (click, choose, make any action). So, you keep scrolling to be sure you are not missing THE best option.

“The virtue is that users want the power and control of the product. But along with that power and control comes that … frustration that can soak up precious watch time: ‘I’m browsing too long and I’d rather actually be watching right now.’”(Glen Davis, Netflix product designer).

“Play Something”: a solution from user research

Netflix solution? The “Play Something” shuffle feature to find something to watch based on previously watched programs to eliminate, or at least reduce, the decision fatigue felt by users. This new shuffle feature is optional: users can select something on the actual home page or click on the “Play Something” button. In this case, the Netflix matrix will choose something to watch and explain briefly this choice while the movie starts.

To launch this new viewing mode, Netflix product team did user research, with concepts in psychology to understand decision fatigue and were inspired by… linear TV.

With linear TV, the decision fatigue is reduced: you just turn on your TV and the program is live. The choice can be difficult because of the multiplication of channels, but users can’t change what is programmed. With Netflix you have both: you can choose or let the matrix chooses for you.

(Note that in France we also have “live content” all day long available on desktop! Like linear-TV with Netflix programs)

Me, in the (French) user sample

I don’t know how user tests have been made in all countries, but as a French Netflix user, I had the opportunity to test this feature. When I clicked on “Play Something” (“Lancer un titre” in French), Netflix recommended me “24” because I’ve watched “La Casa de Papel”. OK, there are series with a lot of action and then? Subjects, contexts, characters are totally different. How do they classify their content? A bit disappointed…

This kind of shuffle feature is appropriate to the music industry with Spotify. When Spotify plays something, you can quickly know if you like it or not, it’s a 3-min song. But with a movie, you can’t decide if you like it or not in less than 3 minutes. Sometimes, users can watch something for a couple of minutes, sometimes for more to give the movie or series a chance. “Keep watching, we don’t know”. And then feel even more frustrated for the time spent!

Another source of solutions: Nudge Theory

To solve decision fatigue on Netflix, the nudge theory can provide different tools to reduce users’ cognitive effort. A nudge has been defined by Thaler and Sunstein (2008) as “a gentle incentive to make the best decision for oneself and society while preserving the freedom of choice of individuals”. The nudge theory has been applied in areas related to the well-being of individuals (health, security, environment…) by governments (Obama in the US, Cameron in the UK…). The nudge theory is based on the concept of bounded rationality, from Simon and taken up by Kahneman (2003) to explain that individuals are biased, lack information, or are drowned in a flood of information to be completely rational in their decision making.

Based on Kahneman’s works, different nudge mechanisms can be divided in (at least) 2 groups: nudges that influence System 1/Automatic and unconscious system when you think fast (1+1 =…), and nudges that influence System 2/Deliberate and conscious system when you think slower (231x4 = …).

In System 1 nudges mechanisms

Changes in choice architecture with salience for example, when you highlight the best option to choose and also default choice in the architecture are considered as System 1 nudges. Indeed, when a choice is pre-selected, users are subject to status quo bias and authority bias (“if someone pre-selected this choice, that should be the best one, I won’t change”). We are not totally rational.

In System 2 nudges mechanisms

The use of social norms including descriptive norms (“my relatives do this”) and prescriptive norms (“the society wants me to do this”) can be useful to design powerful messages (System 2). Researchers have shown that descriptive norms are more effective than prescriptive norms (Bicchieri & Xiao, 2009). In those mechanisms, we have also positive emotions and recommendations.

Nudges to the rescue of Netflix?

On Netflix, the ranking of the top 10 most-watched films in the country, plus the “Play Something” is at the same time System 1 and System 2 nudges to reduce cognitive effort. System 1 nudges because movies are “highlighted” to users, and System 2 because those movies are selected from a “social” ranking, the most-watched movies. But, the most-watched movies?… The ranking is influenced by views, and views are influenced by marketing and promotion. That doesn’t mean I am going to like the same movies like the whole Netflix community (not so speak of judgments on movies but of users’ preferences). The algorithms, at the origins of the Netflix matrix, are too “automatic” to be relevant for users’ preferences.

So how Netflix can reduce decision fatigue with other nudges? Design a new social proof, yes, but a more relevant one with users’ networks and relatives. What all Netflix users like is not necessarily what every user likes. However, a user shares preferences with friends, family, who are (often) other Netflix users. How many times have I stopped my user journey on Netflix to ask friends on Whatsapp: “hey, what’s good to watch right now on Netflix? What do you recommend me?”

What if Netflix became more social? As presented above, the norms followed by loved ones are more effective in influencing an individual, than the norms followed by society at large. By a phenomenon of psychological distance (Liberman, Trope & Stephan, 2007; Trope & Liberman, 2010), I will tend to do, to resemble those who are closest to me in my daily life, those with whom I share commonalities. Those closest to me are far more important opinion leaders to me in near-future preferences than the whole Netflix community, which is far too large and impersonal (Zhao & Xie, 2011). Probably a technical utopia, but definitely something to think about.

Sources :

Bicchieri, C., & Xiao, E. (2009). Do the right thing: but only if others do so. Journal of Behavioral Decision Making, 22(2), 191–208.

Kahneman, D. (2003). Maps of bounded rationality: Psychology for behavioral economics. American economic review, 93(5), 1449–1475.

Liberman N., Trope Y. et Stephan E. (2007), Psychological distance, in E.T. Higgins et A.W. Kruglanski (coord.), Social psychology: handbook of basic principles, New York, Guilford Press, 353–383.

Thaler, R. H., & Sunstein, C. R. (2008). Nudge: improving decisions about health. Wealth, and Happiness, 6.

Trope, Y., & Liberman, N. (2010). Construal-level theory of psychological distance. Psychological review, 117(2), 440.

Zhao, M., & Xie, J. (2011). Effects of social and temporal distance on consumers’ responses to peer recommendations. Journal of Marketing Research, 48(3), 486–496.