Are You Still Watching?
Exploring Unintended User Behaviors
and Dark Patterns on Video Streaming Platforms
Location
Weave Lab @ IIIT-Delhi
Team Members
Akash Chaudhary, Kyzyl Monteiro
Have you ever fallen victim to "just one more episode", stuck in the homepage quicksand or scrolled to infinity to find what to watch? You're not alone.
Video streaming platforms are designed to be captivating, but sometimes their design choices can influence our viewing habits in unintended ways. These subtle design elements, known as dark patterns, can influence our psychology and nudge us towards excessive watching.
This project delves into the world of dark patterns in video streaming, exploring their impact on user behavior and proposing design solutions for a more mindful viewing experience.
As a student UX researcher, I played a key role in:
Designing and conducting a diary study with 22 participants, tracking their moods, emotions, and viewing behaviors.
Leading user interviews to gain deeper insights into their motivations and experiences.
Collaborating with the research team to analyze the qualitative data and identify key themes.
Contributing to the writing and publication of the research paper.
^ Watch the video presentation for DIS'22: ACM SIGCHI Conference on Designing Interactive Systems
This project was part of an extended internship at Weave Lab at Indraprastha Institute of Information Technology, Delhi.
Members of the research team included
Akash Chaudhary (Team Lead) and
Kyzyl Monteiro (Co-author), and were guided by Dr. Aman Parnami and
Dr. Angus Forbes.
Problem
Understand motivations and values
Intrigued by the soaring popularity of video streaming, we embarked on an initial survey to explore the motivations behind this trend. With a focus on university students, we aimed to understand:
1. User Feelings and Motivations: What drives students to watch videos? Do emotions and moods play a role?
2. Content Availability and Influence: How does the ease of access and the power of content narratives impact viewing habits?
3. Social Factors and Purpose: Does watching alone or with others influence behavior? What are the intended purposes of video watching (entertainment, learning, etc.)?
4. Planned vs. Unplanned Behavior: Do students primarily watch with intention, or do they fall into unplanned binges?
Using Pearson correlation analysis on 180 participants, we uncovered valuable insights.
Redder hues in the attached visualization indicate positive correlations, while cooler tones represent inverse relationships.
Key Finding: Habitual behaviors emerged as a significant driver of excessive video watching. This finding served as the springboard for our subsequent research, leading us to develop a deeper understanding of the "habit loop" in the context of video streaming.
Diary Study
22 Participants
228 sessions
Online + Follow-up Interviews
Autoplay
"Five seconds left... never mind, it's started already. I can't stop it now!" (P14)
"Autoplay does not give you timeto think, you are still contemplating about the last episode, andthen the next episode starts." (P3)
Mindless viewing: Autoplay led to a 24.8% increase in mindless viewing compared to 3.5% for recommendations.
Sense of urgency: Users reported feeling a pressure to continue watching due to the Autoplay countdown, often exceeding their intended viewing time.
Lost control: Participants expressed frustration at the lack of agency and regretted succumbing to autoplay's pull.
Recommendations
"Recommendations are great at first, but then... it's like a rabbit hole, and I lose track of time and what I even want to watch." (P8)
Decision fatigue: The constant stream of suggestions overwhelmed users, leading to choice paralysis.
Fear of missing out (FOMO): The pressure to keep scrolling and not miss "better" content fueled impulsive viewing and less mindful engagement.
Regretful choices: Participants often felt they ended up watching content they didn't truly enjoy due to recommendations.
Dual Process Theory: We leveraged this theory to understand the interplay between conscious and subconscious decision-making during video selection and viewing.
PANAS and SAM Scales: To measure user affect and emotional states with precision.
Identifying Patterns
UI Analysis: We analyzed UI elements across 4 platforms, identifying 44 potential dark patterns.
Expert Review: We independently confirmed the appropriateness of identified patterns.
User Interviews: For a user-centered approach, 15 participants (mix of regular viewers and self-reported bingers) were interviewed about their experiences with specific UI elements.
Analyzing Impact
Thematic Coding: We independently coded interview transcripts, identifying 8 themes related to unintended viewing.
Theme Refinement: Themes were converged into 5 distinct categories reflecting dark patterns.
Impact Assessment: Interviews confirmed the hypothesis that dark patterns negatively impact user autonomy and self-control.
Our key findings revealed a disturbing trend: while features like Autoplay and Recommendations offer undeniable ease-of-use, they often come at the expense of user autonomy and control.
Our analysis identified five distinct dark patterns lurking within these platforms:
Click the dark patterns to know more!
Extreme Countdown
Ticking Timers
Insta-play Trailers & Previews
Attention Quicksand
Infinite Scroll
Bias Grind
Feature Fog
Hidden Time Indicators
Hidden time indicators make users lose track of how long they've been watching.
Unplanned binge-watching, exceeding intended viewing time, unawareness of time spent.
Similar to "menu engineering" in restaurants, where expensive items are hidden.
"I tend to forget to keep track of time and it (Netflix) doesn’t help
me at all"
Suggestion
Implement visible "time elapsed" and "time remaining" features for informed viewing.
Extreme Countdown
Ticking Timers
Rapidly ticking timers force users into quick decisions, often leading to unintended binge-watching.
Users feel pressured and lose control, sometimes continuing with content they aren't enjoying due to sunk cost fallacy.
"Autoplay does not give you time to think, you are still contemplating about the last episode..."
Implement customizable timers and easily accessible "off" options for Autoplay to empower user control.
Suggestion
Hidden "Log Out" buttons make it difficult to end sessions, encouraging extended and potentially mindless viewing.
Users feel pressured and lose control, sometimes continuing with content they aren't enjoying due to sunk cost fallacy.
"I would rather just quit the app than go and search for the logout option."
Make "Log Out" easily accessible from the main interface to empower users to end sessions and avoid unintentional binges.
Suggestion
Attention Quicksand
Insta-play Trailers & Previews
Autoplaying GIFs and trailers instantly grab user attention, diverting them from their intended browsing and potentially leading to regretful viewing choices.
Users feel pressured and lose control, sometimes continuing with content they aren't enjoying due to sunk cost fallacy.
" There’s hardly any space on the
screen where I can place my cursor so that there is no animation!"
Implement conscious interaction methods like double-clicks or taps to initiate video previews, offering control and reducing unintended distractions. Additionally, consider mute buttons for audio previews.
Suggestion
Bias Grind
Infinite Scroll
Endless, personalized recommendations based on watch history trap users in "choice paralysis," leading to regretful viewing and potential bias reinforcement.
Users felt overwhelmed and dissatisfied with the endless scroll, often choosing default options or irrelevant content due to decision fatigue.
" YouTube really just provides all the options of what you like, and you feel like you can watch something but are never satisfied."
Limiting the number of recommendations displayed per category.
Requiring additional interactions (like a separate button press) to unlock more options, adding "design friction" to encourage mindful choices.
Suggestion
Our work offers valuable insights for UX professionals and platform designers, advocating for:
Conscious Interaction Design: Implementing deliberate user actions instead of automatic features, empowering viewers to make informed choices.
Transparency and Control: Providing clear and accessible information about viewing time, offering easy-to-find "Log Out" options, and empowering users to customize recommendations.
Promoting Mindful Engagement: Encouraging platform features that foster active participation and conscious content selection rather than passive consumption.
1. Design Interventions for Mindful Engagement:
Friction vs. Frustration: Develop and evaluate UI features that introduce intentional "friction" (e.g., confirmation prompts, time-based delays) to encourage mindful decision-making without hindering usability.
Gamification for Good: Explore gamified elements that promote healthy viewing habits, such as earning rewards for taking breaks or setting time limits.
Personalized Nudges: Leverage user data and preferences to personalize nudges that guide users towards diverse content and discourage excessive binge-watching.
2. Exploring Diverse User Groups and Content:
Live-Streaming Platforms: Analyze live-streaming platforms with unique engagement models (e.g., virtual rewards) and identify potential dark patterns and ethical concerns.
Short-Form vs. Long-Form Content: Compare the prevalence and impact of dark patterns across different content types (e.g., short videos vs. long-form series) to inform platform-specific design interventions.
3. Fostering Collaboration and Advocacy:
Open-Source Design Toolkit: Collaborate with UX professionals to create an open-source toolkit for identifying and mitigating dark patterns across various digital platforms.
Industry Dialogues: Advocate for industry-wide standards and regulations that promote ethical design practices and protect user well-being.
This research journey has been a transformative experience. It deepened my understanding of the ethical complexities within user interface design and the delicate balance between ease-of-use and user autonomy. Witnessing the impact of dark patterns on user behavior solidified my commitment to designing for well-being and promoting responsible user experiences.
The most rewarding aspect has been the process of uncovering hidden design elements and translating them into actionable insights. This research has not only equipped me with valuable skills and knowledge but has also ignited a passion for advocating for ethical design practices within the tech industry.
