Adding a Feature to Netflix

Recommendations from friends and movie critics to solve the dreaded question, “What do I want to watch this time?”

Overview

Background

The term ‘decision fatigue’ has become all too commonplace in today’s world. Whether it’s the number of health insurance providers available or the variety of toothpaste brands to choose from at the supermarket, the feeling of decision fatigue or decision paralysis looms over us everywhere.

The problem is especially bad when it comes to online streaming platforms, a.k.a. deciding what to watch over dinner. The thousands of titles available on Netflix—6,621 to be exact—paired with the modern human brain’s notoriously short attention span means just what you think it does. A whole lot of frustrated people on sofas and an even bigger number of missed opportunities for Netflix.

How can I help users find more relevant and personalised content, reducing the time spent deciding what to watch?

Solution

According to a Nielsen survey, 83% of global consumers trust recommendations from friends and family over those from other sources of advertising. With this in mind, I decided to explore a friend recommendation feature that could help users expedite their decision-making process on the Netflix app.

Goals

Find out what users’ pain points and challenges are around deciding what to watch on Netflix and whether a recommendation feature will help address their needs
Understand where users currently get their recommendations from and which forms of recommendations are the most helpful
Learn how users expect to use Netflix and what their ideal experience looks like
Identify users’ privacy concerns (if any)

Research

What social exchange features do competitors have and where can Netflix improve?

After looking into other streaming apps (Hulu, Prime, Disney+), I immediately realised that no one has been tackling the problem of recommendations. The closest I could find was apps allowing the sharing of content via a link or other social apps.

I decided to look beyond the immediate competition and look into similar apps (YouTube, Spotify, Letterboxd). What I found interesting here was that many of these apps have a playlist or activity feed feature in which users can see what content their friends or following are consuming. However, these types of watchlist features were not at all tailored to the user in the way that recommendations would be.

User interviews

In order to validate my hypothesis that users would be able to reduce their time browsing Netflix titles by being able to receive recommendations from friends within the app, I needed to first speak with real Netflix users and understand their actual needs and pain points.

Participants: 5 users (2 men + 3 women)

Age range: 22-42

Location: UK and Japan

Criteria:

User of Netflix
Uses the app on a mobile device
Finds it difficult to decide what to watch
Highly interested in a friends’ recommendations feature on Netflix

Affinity mapping

What were some of the key insights from speaking to users?

Decision Paralysis

Deciding what to watch on Netflix is a frustrating and time-consuming experience due to the amount of options and lack of personalised recommendations.

Convenient & Seamless

Users want their experience to feel effortless and seamless with the existing UI of Netflix rather than resembling a social media app

Friends’ Recommendations

Friends and families' recommendations are the most relevant and helpful way to help them decide what to watch. Users are motivated by personalisation and intimate interactions between friends.

Movie Critics

Other users are more interested in receiving recommendations from movie critics or people with ‘good taste.’ Many go to other sites to check ratings before coming back to Netflix.

Privacy

Users want to feel secure and in control of their privacy because they are worried about being judged by others if other people have complete access to what they have been watching.

User personas

In speaking to users, I found that there were two camps of people with slightly different needs. There was one group who were highly receptive to and motivated by the idea of sending and receiving recommendations from friends, and another group that was slightly less interested in what their friends are watching and more interested in getting their recommendations from famous movie critics online.

Defining the problem

User flows

As I started thinking about how to conceptualise this feature, besides the actual recommendations, broader questions remained around how to add your friends and how to give users the most precision over their settings to ensure they felt in control of their privacy. In tackling these questions, I ended up with a total of four user flows:

Add friends and follow movie critics
Send recommendations
Browse recommendations
Settings

Usability Testing

Once my low-fidelity wireframes were ready, I recruited and screened five users to test with who matched the two target personas. I tested all four task flows (pictured above) and was focused on receiving feedback on the flows themselves rather than the UI elements, seeing as I was going to be utilising Netflix’s existing design system instead of my own branding.

The usability testing sessions were immensely helpful in helping me better understand users’ mental models, especially in regards to navigation and language. All users were able to complete the tasks within a reasonable time, but the feedback around the Recommendations Preferences flow was the most pronounced and I knew that was where I had to prioritise my iteration efforts.

Iterations

Final Prototype

To let people send each other recommendations, first we need to let them add friends.

Users can send their friends recommendations via a monthly pop-up that shows their recent titles.

Recommendations help users connect more deeply with friends and most importantly, make quicker and better decisions.

If users want to discover & follow famous critics, then there needs to be a place where they can see who they’ve followed.

Users have control and precision over what they see and have the option to change their preferences at any time.

Conclusion

All in all, this project was a new and exciting experience as it challenged me to channel my creativity within an existing design framework. Compared to the branding I had done for previous projects, this project pushed me to think up novel ways of utilising Netflix’s own design system to speak to user needs and goals.

I enjoyed interviewing users, validating my hypotheses, and exploring the various design solutions that could address their challenges and pain points. However, more exciting than validating what I had already assumed was discovering new insights, namely that movie critics–something I hadn’t even considered–were also a very sought after source of recommendations.