PROJECT BRIEF AND PROBLEM STATEMENT
Netflix as a platform is amazing for the amount of content it has for its users. The consumer is able to select from over 3500 films and TV-Shows which is awesome for having a wide range of genres and films to select from. This in itself is where the problem comes in. I’m sure you have experienced it yourself or have been in the room while others have gone through it, and that is, the endless scrolling of selecting what to watch. You have definitely heard the joke that says that users spend more time deciding what to watch than actually watching their movie. Why does this happen and how can we use a UX approach to reduce the time, and even perhaps improve the experience in deciding what to watch?
CONTEXT
I was tempted to ask myself, “How can we use a UX approach to reduce and eliminate this problem?”, but to be honest I’m not sure it can be completely eliminated since we are humans who struggle at making decisions when many equivalent choices are available to choose from. In this case, Netflix provides thousands of films, reducing the number of movies and shows will only lead to a decrease in the number of users. This can be attributed to modern technology as we have access to more information, products, opportunities and overall range of emotions.
It is mentally draining for us to select a single option because each option must be weighed against all other options in order to select the one that will give us the best experience. The issue which we are here to focus on and goes hand in hand with the problem above, is how much time we spend scrolling to find the correct movie that matches our mood or preference. It has been reported that we spend about 18 minutes on any given day just scrolling trying to find what to watch. It surely sounds like a first world problem; we are offered thousands of options, yet they’re all too good for us to choose a single one, and how do we go about organizing all the content? We become overwhelmed with the number of options available that we don’t take pleasure when we finally decide on the one. We ultimately end up feeling the opposite of what we wanted to do when we first sat down on the couch, and we might even end up not watching anything because of the number of options available. We are constantly in pursuit of the ‘perfect’ choice for us to feel accomplished.
USER RESEARCH
I began by collecting some user data and creating a survey to understand the viewing habits of Netflix Users. I wanted to understand how much people say they spend selecting a movie or TV-Show prior to watching, if they believe Netflix provides correct recommendations for similar content to watch, what they believe the best way to rate content is, how they feel about the content they watch once they view it, how they select what to watch, and also what they do if they can’t decide what to watch.
With the survey, I had 77 participants most between the age range of 18–25 who provided responses and the opportunity to learn who our users are, what they want to accomplish, what information is useful to those users and where the users are struggling the most. It provided quantitative data and gave us our ‘what is happening?’. Without a good understanding of users’ current paint points, we won’t be able to create goals for product improvements.
The survey and results can be seen here.
Taking our quantitative data that we received, we use those data points to build metrics, work out what feature of the service we want to focus on and build an engaging persona which identifies the user goals, their preferences and also the frustrations and pain points when using the application. The survey gave us our ‘What is happening?’, we now process the results that will give us an answer to ‘Why is it happening?’
WHAT DID WE LEARN BASED ON RESEARCH?
1. Users spend around 2–4 hours a week on Netflix and about 10–15 minutes browsing for content prior to selecting something to watch.
2. Most users go into Netflix not knowing what they want to watch.
3. Users spend more time choosing something to watch when being in a group of people than if they were on their own.
4. There is a grey area between loving and hating a movie (thumbs up & thumbs down). Users aren’t able to choose that middle ground.
5. Users trust other user reviews when selecting what to watch. They also trust friend recommendations.
6. Users do watch content more than once.
7. Users choose content based on their mood.
8. Most feel stressed when they can’t decide what to watch.
9. Most users completely turn off Netflix when they can’t decide what to watch. (important)
DEFINING OUR IDEAL USER
SETTING GOALS
As a UX approach to this problem, and having an understanding of the insights and frustrations users go through when watching Netflix, based on our findings:
If we created an enhanced Netflix discoverability application,
it will solve for improved movie and TV-Show selection times with a considerable growth in satisfaction of the content being consumed,
leading to more user engagement and greater trust in Netflix recommendations.
Based on the frustrations we discovered, we want to enhance the experience by looking at solving the current issues. These frustrations become the user needs and we turn those needs into opportunities for improvements. We can see that many of the current issues are preventable, which we should take into consideration in our possible solutions. Our overarching goal is to make the endless scroll feel a little less endless.
EXPLORING SOLUTIONS
1. Improve rating system. Use the star rating rather than a thumbs up and thumbs down rating system to help guide in decision making when selecting a film.
2. Separate recently watched. Hide the movies and TV-Shows on a separate page so users don’t have to scroll through those already seen. - users have to do more searching
3. Randomize a Movie. When users are unsure of what to choose, Netflix will randomly select something to watch based on their viewing history.
4. Show popular/trending films. Create a category which showcases only trending content.
5. Connect with Friends. It was proven that users watch shows and movies based on friend recommendations so this may be useful for keeping users locked into Netflix for longer.
6. Organizing films by the mood. Alongside the genres filter, it may be possible to organize content based on the mood that is experienced after watching the film.
PRIORITIZE AND SELECT BEST SOLUTIONS
The next step is to plot these solutions on an Impact/Effort Matrix Chart to assess their feasibility. We want to pick solutions with an overall high impact and low effort.
BEST SOLUTIONS
On Improving movie rating system (#1):
For one, close to 70% of the participants believed a movie should be rated using a star rating system followed by the next popular option at 15% which was by the quality of comments left by other users. The current system Netflix uses is giving a thumbs up or down to recommend movies in the category but upon doing research online, many users insisted this system is flawed because it puts the rating at two extremes, you either hate the movie or you love it, there is no in-between. Although this may work at a personal level, the user has no idea how others feel about the movie. This forces them to pick up their phone and search information online about the film, in which case the user is losing retention with Netflix.
The star rating system I suggest is not new and is widely used on many other websites and applications which users are familiar with. It represents the average rating of all the users who watched the film and guides the user before making a decision. Works in both directions, users can look at reviews and also write their own review. — To add, I might be getting a little technical here, and you may think, what happens if someone simply decides to leave a bad review in hopes of bringing the ratings down? As a way to prevent this, ratings can only be left at the end once the movie has been seen. It makes sense because why would a bad review be left when the movie has not been watched. In regard to comments, since most people are using the TV remote, typing a long sentence may be difficult, possible solutions can be AI generated words or phrases that can be selected to leave a review which users can select from. We learned this from the survey completed, that users do trust other user generated reviews which they use as filters to decide what to watch.
On evaluating content:
The problem with video streaming services like YouTube and Vimeo is that we can’t assess the quality of the video until we actually experience it. It’s no secret that users prefer quality over quantity, what’s interesting is that Netflix in a way forces the average user to search through everything in order to teach Netflix what quality content means for each user, and that way helps at bringing more shows on the platform which users might be interested in. There’s a drawback here though, instead of users having to constantly provide feedback to Netflix to teach the AI what we may like to watch — which in my opinion isn’t always correct based on countless reviews left online regarding this subject — the star rating system would work in the opposite way in that users help other users provide similar content. It’s a more human feel to the entire system.
This rating system is great for informing and guiding decision-making, and ultimately providing guidance on how others feel about the content, which as we also learned many users trust in.
On connecting with friends (#5):
A very high percentage of all apps available to download now-a-days include some sort of ‘share with friends’ or ‘connect with friends’ option. It’s clear to see people like to interact with those around them and take into consideration their opinion when making decisions. Simply taking a look at the survey results, over 88% of the participants stated they watch content based on friends suggested recommendations. Psychology also teaches us that there are several reasons as to why people feel inclined to share content online. First because they want to better the lives of others, second, they want the content to reflect their online identity, third because they like the feeling of having others comment and engage with the content, and lastly because people want to spread the word about something that they believe in. These are all very key solid points for adding a social aspect to the application, and to put it to simplest terms, it gives people something to talk about therefore creating a stronger emotional reason to share the content. It’s the same reason why posts go viral on social media, because they have higher emotional value to the subscriber. I believe this option would pair great with Netflix.
But how would people connect with other users? First it goes back to our previous point on how people evaluate and rate content to let others know what they think about the movie or show, second, users can add friends to their account and see the content their friends are watching to get recommendations, begin a show at the same time as their friends, as well as see what their favorite shows and movies are.
On selecting content based on the mood(#6):
A point that struck me as being key in our redesign which was confirmed on the survey results was that users choose content based on the mood they are feeling at the time of watching. This was confirmed by 92% of the participants and I would also count myself in that group. We don’t consume the same content every time we log in, it differs between parameters such as time, day of week, location, device and many other factors. I asked myself, what’s the first thing I do when I sit down to watch a movie? And I’m sure you do the same, and that’s to ask yourself, “What am I in the mood for?”. I feel like we do this as a way to talk ourselves into figuring out how we want to feel after watching the movie, also to understand what kind of emotions we want to experience by the end of the show. To do this, we can explore this solution by adding a tab where users select not by genre but rather by mood.
WIREFRAMES BASED ON BEST POSSIBLE RESULTS
Star Rating:
Viewing friend recommendations:
Selecting movie based on mood:
FINAL RENDERS
Star Rating:
Viewing friend recommendations:
Selecting movie based on mood:
MEASURING SUCCESS
In order to define if the solution has been achieved, we need to have defined our KPIs that will give us a measure demonstrating how effective it was at solving our goals. A successful solution to this problem can be quantified by first, noting down if users are able to take advantage of the different tabs available to choose from to aid them in selecting the content to watch. Second and most importantly we would test users on how long they spend selecting a movie based on the posed recommended solutions. We want to reduce the time and number of decisions the user has to take before selecting the movie or show, provide the solution which in this case provides the most satisfaction, desirability and value to the viewer as well as have a greater conversion for users to want to watch content once they have turned on Netflix rather than walking away.
PUTTING IT ALL TOGETHER
You can view the working prototype here.
USABILITY TESTING
Given the chance to perform user testing, the iterative process would be as such:
- Asking users to try selecting a movie or show and see if they notice the different tabs and options on the screen available to help them select the content. This would be the first important point.
- Second, make note of difficulties and pain points for further improvements. Understand what works and what doesn’t.
- Lastly, collect feedback from users on ease of navigation for the application.
The tasks that would be asked from each user would be to (1) explore the rating tab for the movie, (2) rate a movie after having watched it, (3) connect with a friend to see their recommendations, (4) select a movie based on their current mood.
NEXT STEPS
It’s important to note that in this personal project I wasn’t looking to redesign the entire user interface of Netflix because I’m sure they have had much more research into doing so, and this isn’t what the project is about, but rather improve on features which would bring a greater positive experience on the platform and get users to spend less time watching just to watch.
I also only explored these solutions on the desktop version to get a better sense of how these features would function. It would also be important explore mobile, table, and TV versions of the updates.
COMPLETE CASE STUDY
I wrote an article on Medium with lots more detail about this case study. You can find it here.
