Netflix (& Recommend)
Designing a social function for those that binge.
Click above to see a video walk-through of the native app!
My team was tasked with finding the source of what drove people to the shows they do end up watching, and how Netflix can create a feature to expand on that. The feature we designed was also meant to be responsive across multiple platforms.
After user research, data analysis, usability testing, and iteration, my team came to the conclusion that Netflix users care about what their friends watch. Therefore, when we found Netflix’s hidden “Recommend” function in their native app, we chose to bring it to the forefront and also incorporated it into their web platform.
Competitor and Company Research
Our team researched streaming services, video content websites (like YouTube), video apps (like Vine), and even some entertainment apps (like Spotify) to create a list of features. We then took those features and compared them among Netflix's direct competitors, being other streaming services. This information allowed us to find where Netflix differed from other services, which helped us come up with survey and interview questions to find if those features (or entirely new ones) were needed.
Survey and Data Analysis
We sent out our survey (which you can view here) to find out about viewer demographics and their behavior. Our respondents ranked what their friends watch as their most important factor when picking a show. We also found that respondents were uninterested in a sharing or social feature, which would highly define our MVP later.
Creating the persona was illuminating as our team found that most behavior was correlated even when you took into account the age, sex, location, and job of the respondent. Therefore, we did not assign a gender, age, or location to our persona, but rather a set of behaviors. Even Netflix’s VP of Product Innovation, Todd Yellin, has announced that Netflix is more or less disregarding “almost useless” data on demographics as it doesn't correlate to user behavior or what shows they choose to watch.
We came up with a list of ideas for features to make or iterate on from our competitive research and general brainstorming. We then were able to eliminate some (like social sharing and content creation) from our survey results. Some features were supported from the survey data, like seeing content based on your friends or just having more content in general. We did four in-depth user interviews and tallied how many of them supported or rejected an idea to come find the three that had the most potential.
After diluting the possible features down to three, we chose to move forward with creating a feature that allowed users to receive content based on what their friends watch. After digging through the native app, we found that there was a "Recommend" feature that people were not aware of, as well as an existing notification system that rarely was used. We chose to move the "Recommend" feature to a more prominent and natural location, create new screens for "My List" and a new "Recommended" list, and use the existing notification framework to onboard and alert users of these new features.
We were tasked with making our feature responsive. Since the app already had some of the framework of recommending, we looked to how the website could implement it. We made a paper prototype putting the button next to the "Add to My List" button that already exists when a user hovers over a title card.
Unlike the website, the Netflix app doesn't have a separate area to view "My List". Instead, it simply has a scrolling carousel for all the titles a user has added to their list. Our users found it annoying to keep scrolling and said they would abandon the carousel for the search bar, which made us come up with the idea of having a separate screen for "My List". This also would allow the addition of a "Recommended" screen for managing recommendations sent by friends.
For the web, in our first prototype we had a small space for people to recommend to their friends and during user testing it proved to be too cluttered as well as visually jarring. We iterated and now when users click on recommend, it expands below to give enough space to navigate.
For the native app, our first iteration of receiving a recommendation was usable but had some issues. It wasn't clear to users that this was a hero image and there were more recommendations when you exited out, the red "X" button looked like a delete button, and writing out the name of who recommended a title would be problematic if more than one person would recommend a title.
We made changes to have the Recommended screen be cards that were stacked and you could scroll through. Netflix already integrates with Facebook so we just used profile pictures to indicate who recommended the title. Also, users simply slide the title to reveal the options to accept or pass on a recommendation; this function was discovered by all users without prompting.
new user flows
Sending a Recommendation
Creating a feature on the web to recommend along with developing screens for "My List" and "Recommended" on the app has many business benefits for Netflix. Financially, there is an incentive for users to keep their subscription if they have a "My List" full of programs they haven't watched yet. Since their friends would be recommending what they watch, and that is the key factor of someone picking a show, they are likely to add those titles and stay subscribed to finish those shows.
Netflix also could use this data and information when making programming decisions. Netflix may be able to prioritize titles based on recommendations you are accepting over what you are passing on. Netflix also may be able to see what shows are highly shareable within their library and acquire more shows like it or develop their own original content that is similar.
Sketch 3 | Axure | Balsamiq | Invision | Marvel | Omnigraffle | Google Surveys | Excel PivotTables | Keynote | Paper Prototyping
Company and competitor research; user survey; user research; data analysis with PivotTables to find correlations; user flows; lo- to hi-fidelity mockups in Sketch3; business benefits; presentation.