Executive Summary

This case study outlines a 3-month research project focused on addressing user frustrations with Netflix's content discovery and recommendation platform. The study aimed to improve user experience, address pain points, and ultimately enhance content recommendation accuracy.

Executive Summary

This case study outlines a 3-month research project focused on addressing user frustrations with Netflix's content discovery and recommendation platform. The study aimed to improve user experience, address pain points, and ultimately enhance content recommendation accuracy.

Problem: Users express widespread dissatisfaction with Netflix's content recommendations and struggle to find shows and movies that align with their preferences.

Solution: Design and implement new content discovery and recommendation features within the next 6 months. Leverage machine learning and data analysis to deliver more personalized and targeted content suggestions.

Key Activities:

  • Reddit Analysis: Examined community forums to pinpoint common user frustrations with Netflix's content selection and recommendation features.
  • Diary Study: Conducted a 4-week diary study with 25 Netflix users to gain in-depth insight into usage patterns, triggers, and pain points related to content discovery.
  • Data Analysis: Utilized affinity diagramming to cluster key pain points and themes extracted from user diaries.
  • Journey Mapping & Task Analysis: Developed a user journey map and conducted a task analysis to visualize the content discovery process and identify areas for improvement.
  • Lo-Fi Wireframing: Created initial wireframes incorporating potential solutions to streamline content discovery within the existing Roku Netflix interface.

Outcomes:

  • Key user needs: Identified critical user desires for increased control over viewing habits and mood-based recommendation options.
  • Pain Point Identification: Discovered core pain points regarding the platform's recommendation limitations and insufficient content search control.
  • Solution Ideation: Generated solutions focused on enhanced search functionality and a potential mood-based recommendation system (contingent on further testing).

Challenges:

  • A/B Testing Limitations: Encountered obstacles in setting up proper A/B testing on an adequate sample size.

Conclusion:

This research successfully uncovered significant user pain points and generated actionable insights for future development. Ongoing user testing and development efforts will guide refinement and implementation of proposed solutions.

Problem: Users express widespread dissatisfaction with Netflix's content recommendations and struggle to find shows and movies that align with their preferences.
Solution:
Design and implement new content discovery and recommendation features within the next 6 months. Leverage machine learning and data analysis to deliver more personalized and targeted content suggestions.
Key Activities:
  • Reddit Analysis: Examined community forums to pinpoint common user frustrations with Netflix's content selection and recommendation features.
  • Diary Study: Conducted a 4-week diary study with 25 Netflix users to gain in-depth insight into usage patterns, triggers, and pain points related to content discovery.
  • Data Analysis: Utilized affinity diagramming to cluster key pain points and themes extracted from user diaries.
  • Journey Mapping & Task Analysis: Developed a user journey map and conducted a task analysis to visualize the content discovery process and identify areas for improvement.
  • Lo-Fi Wireframing: Created initial wireframes incorporating potential solutions to streamline content discovery within the existing Roku Netflix interface.
Outcomes:
  • Key user needs: Identified critical user desires for increased control over viewing habits and mood-based recommendation options.
  • Pain Point Identification: Discovered core pain points regarding the platform's recommendation limitations and insufficient content search control.
  • Solution Ideation: Generated solutions focused on enhanced search functionality and a potential mood-based recommendation system (contingent on further testing).
Challenges:
  • A/B Testing Limitations: Encountered obstacles in setting up proper A/B testing on an adequate sample size.
Conclusion: This research successfully uncovered significant user pain points and generated actionable insights for future development. Ongoing user testing and development efforts will guide refinement and implementation of proposed solutions.

The Business Problem

Over the last three months, I've encountered numerous users expressing their discontent with the Netflix streaming service, particularly in its recommendations that often fail to align with their preferences. While there are additional issues I've observed, the overarching concern is users' dissatisfaction with the content they desire and are actively seeking. My motivation for delving into this stems from a desire to showcase my comprehensive research skills, understanding of user behavior, adeptness in identifying pain points, and the capacity to propose potential solutions that would not only appease users but also align with the goals of the organization.
Long Term Goal: Design and develop a content discovery and recommendation platform in the next 6 months that helps Netflix users discover new content based on their preferences and viewing history. The platform should leverage machine learning algorithms & data analytics to provide personalized recommendations and enhance the user experience.
One of the biggest concerns I had about the outcomes of the results is would the new features not help find relevant results to the user's interest?

The Approach

Problem Identification:
My research began with exploring online forums like Reddit to investigate common user frustrations with the Netflix platform. Key complaints centered on difficulty finding specific content, irrelevant recommendations, and general dissatisfaction with Netflix's focus on pushing its own agenda rather than catering to individual preferences.
While other issues arise in the forum post such as wanting to have the ability for the movie never show up again if already watched it, and list all movies or shows from A to Z I wanted to dive deeper to see if their were any other issues that users were facing when it comes to Netflix search.
User-Centric Research:
To deeply understand user behavior and pain points, I designed a 4-week diary study. I carefully recruited 25 Netflix users from Charlotte, NC through a participant survey, selecting individuals with varied usage habits and engagement levels to identify a range of potential issues.

Survey Insights:
Survey questions were strategically formulated to reveal pain points and potential triggers. Results pointed toward users favoring targeted searches and browsing over reliance on Netflix recommendations, emphasizing dissatisfaction with the platform's understanding of their preferences.
I use this survey question to help select which participants to have in my diary study by using people that fall under the every day and several times a week category. I use those categories to have participants that knows the Netflix platform and shows that they care about it since they use it most often than the other categories.
These 2 questions were designed to test participants with a pain point and trigger question to give a glimpse to see the current state of how they felt finding new content and searching for precise genre of series or movies content on Netflix. From the results in the new content question shows that they prefer searching for specific titles and browsing categories than relying on Netflix's recommendations which points to the main problem of not understanding user's preferences. The second question shows that users need changes that help users find precise movies/series.
The inclusion of the question about the least favorite aspects of Netflix's search bar in the case study is essential for understanding user sentiments and improving the overall user experience. User feedback on the lack of relevance in search results guides enhancements to our content recommendation algorithms, ensuring that users receive content aligned with their interests. Complaints about the difficulty in finding desired content inform improvements in navigation and user interface design, making the search process more intuitive. This user-centric approach not only enhances satisfaction but also positions us competitively in the dynamic streaming landscape. The  reliance on ratings for quality assurance, these insights guide our efforts in delivering a personalized and satisfying content experience, ultimately enhancing user engagement and loyalty.
2nd survey question regarding the most important factors when choosing a movie or TV show in our Netflix search case study is fundamental to shaping our content strategy in line with user preferences. By understanding users' priorities among factors such as genre, actor, director, and ratings, we gain valuable insights into the key elements that influence their viewing decisions. This data serves as a foundation for optimizing our Netflix search and recommendation algorithms, enabling us to prioritize and curate a catalog that aligns closely with what our users value most. Whether it's the allure of a specific genre, the presence of a favorite actor or director, or the  reliance on ratings for quality assurance, these insights guide our efforts in delivering a personalized and satisfying content experience, ultimately enhancing user engagement and loyalty.
This survey question helps me identify a pain point with users by understanding that users has difficulty in finding results that are relevant to their interest. Learning about this shows users are facing problems with their results not relevant which I need to dig deeper on what caused it.  
Diary Study Setup:
Over four weeks, I engaged with participants to produce over 500 in-depth diary entries. These open-ended prompts focused on experiences and triggers relevant to Netflix search. To manage this substantial dataset, I developed a detailed coding scheme covering aspects like search results, personalized recommendations, genres, emotions, technical issues, and more.
In the coding scheme I focused on the following sections to cover all aspects of Netflix search such as ratings & reviews, privacy concerns, frustration with browsing, mood and emotions, Netflix My List, Inaccuracy of recommendations, Movies vs TV series, technical issues, tags and labels, overall content discovery experience, pain points with search results, challenges in finding content, continuing to watch, genre and theme, trailers and teasers, external recommendations, actors creators and directors, foreign language content, initial thoughts, personalized recommendations, trending now and top picks, thumbnails and descriptions, outside preferences, because you watched, and unexpected content.
Over the past four weeks, the sheer volume of diary entries I've been receiving has threatened to overwhelm me. To maintain my sanity and keep track of these entries, I've devised a system of coding schemes that has proven to be invaluable. By assigning unique codes to different categories of entries, I'm able to quickly identify and prioritize entries based on their importance and urgency. This has saved me countless hours of sifting through endless pages of text, allowing me to focus on the most critical information. Additionally, the coding system has helped me to identify patterns and trends in the entries, providing valuable insights into the overall sentiment and experiences of the individuals submitting them. As the volume of entries continues to grow, I'm confident that my coding schemes will remain an essential tool for managing this influx of information.
When it came to creating the questions I focused on giving the participants open ended questions that doesn't allow them to say yes or no answers. I also encourage the participants throughout the 4 weeks to fully express their thoughts, feelings, and their experiences about using the platform daily. At first they only produce 2-3 sentences during the first week but after giving them encouragement and feedback on their entries the next 3 weeks went to produce 3- 5 sentences to the point of even making paragraphs of their experience.
Communication and Engagement:
Regular email communication with participants was vital to maintaining transparency, clarifying any issues, and encouraging detailed feedback. Their entries improved significantly over time due to active support and encouragement.
To gain a comprehensive understanding of the participants' experiences and identify recurring themes, I decided to transform my own room into a visual representation of the diary entries. This immersive approach allowed me to physically surround myself with the participants' words and emotions, creating a powerful and impactful medium for analysis. By spreading the entries across the walls, I was able to simultaneously view the entirety of their experiences, gaining valuable insights into the overall sentiment and common threads that emerged from their daily interactions with the platform. The physical layout of the entries facilitated pattern recognition, enabling me to identify recurring issues, concerns, and positive aspects of their experiences. This unique visualization technique provided a holistic perspective on the participants' journey, allowing me to delve deeper into their lived experiences and the impact of the platform on their daily lives.
Immersed in a sea of diary entries, my room transformed into a sanctuary of understanding, where the participants' voices echoed through the walls. Each entry, a fragment of their lived experience, became a piece of a puzzle, slowly revealing the intricate tapestry of their journey with the platform. The physical proximity to their words allowed me to absorb their emotions, their joys, and their struggles, creating an empathy that transcended the limitations of text.
Immersive Data Analysis:
To visualize findings and patterns, I transformed my room into a physical display of diary entries. Surrounding myself with user comments allowed me to identify key themes and gain deep insights into their emotional impact.

The Results

Key User Challenges

Limited Search & Filtering:
Users report frustration with limited options for filtering and searching, leading to repetitive results and difficulties finding niche content. Desired improvements include filters for year, genre relevance, director, user ratings, and run time.
Content Organization:
Issues with visual design, layout, and thumbnail appearance make browsing difficult. Larger thumbnails with information on hover would add clarity.
Inaccurate Recommendations:
The search algorithm often misunderstands preferences, leading to wasted time and irrelevant suggestions. Users want keywords and content variety that better reflects their interests.
Solutions Informed by User Data Content Discovery:
  • Enhance search within TV series to find episodes
  • Organize and refine the watchlist feature
  • Consistent content display across devices
  • Improve tracking of watch history
Decision-Making Tools:
  • Integrate IMDB-like information
  • Address technical issues via user feedback
  • Add mood-based filters
  • Improve overall navigation
Privacy & Customization:
  • Provide transparent data controls
  • Allow genre organization
  • Add sorting options, improve trailers
  • Integrate external recommendations
  • Personalize filters (actors, directors, creators)
Foreign Language Content:
  • Prioritize preferred subtitles/translations
  • Ensure subtitle availability across languages
  • Offer regional genre filters
  • Improve subtitle accuracy and sync
  • Add subtitles for less common dialects
  • Enlist expert/user help for subtitle refinement
Additional Considerations:
  • Use ratings/reviews to reflect mood
  • Let external recommendations guide exploration
  • Focus on creating niche-specific filters
  • Improve recommendation algorithm accuracy
The Foreign Language filters content offers a comprehensive approach to enhancing user experience. It includes features such as customized language selections with a priority on preferred subtitles or translations, ensuring the availability of all subtitles, and introducing regional genre filters. To address accessibility and accuracy, efforts are made to import subtitle accuracy, provide subtitles for less common languages and dialects, and enlist user or expert assistance in improving subtitles and addressing audio sync issues. Initial considerations involve incorporating mood into decision-making through ratings and reviews, while external influence filters allow users to explore different perspectives through outside recommendations. The emphasis on exploring niches is achieved through customized filters. Personalized recommendations focus on aligning algorithms with user preferences, allowing users to indicate content preferences for future recommendations. The overall goal is to create an enjoyable content experience, with continuous improvements aimed at satisfying users.
Ratings and reviews play a crucial role in influencing user decisions, ensuring that user-generated feedback contributes significantly to the decision-making process. The incorporation of trending now and top picks features is aimed at enhancing user influence, with a focus on creating filters that optimize these categories. Customizable filters and recommendations further personalize the user experience by displaying new trending content and top picks based on individual viewing habits. Thumbnails and descriptions are streamlined to provide concise and accurate reflections of movies and shows from their artwork. External recommendations are improved by refining algorithms based on movie ratings, new releases, and user viewing patterns. Additionally, reactions to personalized recommendations are enriched by incorporating mood and curiosity in the previews. Efforts are made to address content issues, ensuring that users can easily find specific content without it feeling time-consuming.
User journey maps and task analysis were created to analyze the Netflix content discovery experience. These tools highlight the decision-making, emotions, and outside factors that shape how users interact with the platform. The task analysis, informed by a diary study, examines challenges users face while searching for content, such as issues with search suggestions, misaligned content, and difficulties finding episodes. Challenges encountered during the project itself included difficulty crafting questions for participants that would reveal pain points, as well as limitations in conducting A/B testing due to insufficient user traffic.
In this comprehensive set of user journey maps within the Netflix ecosystem, various scenarios depict the diverse experiences of viewers during content discovery sessions. Whether exploring personalized recommendations, venturing beyond usual preferences, searching for specific themes, encountering unexpected content, returning to incomplete series, relying on ratings and reviews, or deciding between movies and TV series, the common thread is the user's dynamic interaction with the platform. The summaries encapsulate the nuanced decision-making processes, emotional responses, and external influences that shape the Netflix viewing experience. From the initial trigger to the final moments of content immersion, each journey highlights the platform's efforts to cater to individual preferences, foster open-minded exploration, and provide an enjoyable space for diverse content discovery.
This task analysis delves into various aspects of the user experience when finding content on Netflix, as revealed through a diary study. The breakdown of tasks, subtasks, and associated challenges provides a comprehensive overview of the user journey. From searching for specific genres or themes to encountering content outside of preferences, refining searches, assessing content organization, dealing with unavailable content, and managing expectations around intriguing titles or thumbnails, each analysis offers insights into user behaviors, challenges, and potential solutions. It also explores how issues with Netflix's search suggestions, content misalignment with expectations, and difficulties finding specific episodes impact user trust and satisfaction. These analyses collectively contribute valuable information for enhancing the content discovery process on Netflix, addressing user frustrations, and improving the platform's overall usability.
Problems that I had encountered when conducting the project was providing the participants the right open questions to cause triggers and pain points, and trying to conduct A/B testing. When first giving the participants trigger questions they responded I was getting responses that shows their natural behavior when navigating the platform, but I couldn't cause any pain points to happen to notice the changes they wanted. When I gave participants more questions that focus on certain issues when they stumble upon it I received better responses that helped narrow down problems that causes pain points to the participants.

The Impact

Going back to the business plan of users being frustrated of having less control in searching content and content not understanding user's taste helped shine on more deeper issues. Having the diary studies be able to bring up more issues that users are going through shows how great I am as an upcoming designer. Currently the project is heading in the right direction with being prepared to test the findings I discovered but still glad that I came this far.
Utilizing insights derived from the affinity diagram, I meticulously crafted low-fidelity wireframes proposing potential solutions seamlessly integrated into the current UI iteration of the Roku Netflix streaming app. My intention was to avoid introducing an entirely novel set of ideas, as I aimed for the new features to seamlessly harmonize with the existing interface, ensuring users experience minimal disruption and do not need to undergo a significant learning curve.
I carefully transformed the initial low-fidelity wireframes into high-fidelity versions, meticulously integrating user feedback to refine the visual concepts. This iterative process illustrates the potential of these ideas to evolve into solutions that genuinely resonate with users and drive meaningful business results.

Other Projects

Graphic Design
You Are The Key Patient Booklet
Product Design
SkyScripted Product Design Case Study
Brand Packaging Redesign
Amazon Go's Food Packaging Redesign