DUBBLE
An ADHD-friendly body-doubling companion
Overview
Dubble is a body-doubling* web application designed to help neurodivergent students overcome task-initiation paralysis. Developed over 9 weeks using inclusive design sprint in HCDE 315 at the University of Washington, our team partnered directly with a co-designer with ADHD* to build a tool grounded in lived experience rather than stereotypes.
*a user who helps decide the direction of design. Every flow was co-created to directly address the emotional and executive barriers that make starting tasks daunting.
Role
UX Researcher & Interaction Designer (Team of 4)
Core Ownership
Task Creation Architecture, Qualitative Coding of Feedback, Step Breakdown Interaction & Cross-Visual Coherence
Timeline
March 2026 — June 2026 (9 weeks)
Tools & Methods
Co-Design Workshops, Qualitative Coding, Paper Wireframing, Figma Prototyping
The Challenge
When tasks feel overwhelming, students with ADHD frequently experience severe executive dysfunction. Starting early is nearly impossible because focus often only activates under the pressure of deadlines, while falling behind triggers an acute cycle of avoidance and self-shame. Existing tools make this worse by relying on rigid countdown timers, gamified distractions, and guilt-inducing broken streaks.
The Solution
Dubble breaks the initiation freeze by automatically generating bite-sized, preset steps from imported online assignments, letting users edit them in one click before starting. Once launched, users focus alongside an ambient, customizable virtual body double (their avatar) at the bottom of their screen. If inactivity occurs, pop-able bubbles float across the screen with self-care nudges, offering neutral accountability to reduce guilt-tripping.
My Role
As the UX Research and Interaction Designer for a four-person team, I personally drove:
Task Creation & Step-Breakdown System: Architected the screen that takes daunting assignments and breaks them into manageable steps using automated presets and LMS imports (Canvas/Google).
Qualitative Synthesis & Research: Co-led in-depth interviews, conducted qualitative coding to organize insights into 4 core behavioral themes, and ran participatory co-design sessions.
System Coherence: Maintained Figma component libraries, typography tokens, and layout consistency across the team’s wireframes and prototypes.
Constraints: A fast 9-week cycle requiring realistic, high-fidelity Figma component logic to simulate live ambient companion states and floating overlay interactions without an active production backend.
Meet the Co-Designer
To ground Dubble in lived experience rather than clinical assumptions, our team partnered with a college student diagnosed with ADHD who actively struggles with executive dysfunction, task initiation, and academic stress.
Rather than treating them as a passive test subject, we brought them in as an equal creative partner for 9 weeks:
Continuous Collaboration: Our co-designer participated in 1–2 hour weekly co-design sessions, interactive brainstorming activities, and critiques.
Direct Feedback Loops: We contacted them between iterations to stress-test early paper sketches, critique our user flow and UI layouts, and evaluate feature trade-offs.
Guiding Voice: Their direct insights into the panic of deadlines, the guilt of falling behind, and their personal dislike of gamified apps directly shaped our core design principles and feature decisions.
The Problem
Existing productivity tools treat users like machines: rigid checklists, countdown timers, and punishing broken-streak notifications. For people with ADHD, these patterns trigger a vicious cycle of executive dysfunction:
The Problem Statement: How might we support students with ADHD in reflecting on mental barriers preventing productivity so that they can find motivation in starting tasks?
Market Analysis: Deconstructing Focus Friend
To understand the current market, we audited existing tools focused on concentration, and I specifically evaluated Focus Friend—a gamified focus timer developed by John Green* that uses a cute avatar companion designed for people who need to concentrate on tasks.
Where it succeeds: It is effective at keeping users motivated to stay focused because users like the avatar style, cuteness, and gamification. It attempts to body-double with the user by showing the avatar actively focused alongside them.
Where it fails:
Fails to address task initiation: It only functions as a timer. It assumes users will suddenly get on task if they don’t use their phone, failing to consider that users can still get distracted by physical tasks around them. It trusts that users will be focused, but gives zero guidance on how to start.
Hidden shame mechanics: Despite marketing itself as a “no-shame” app, it still uses shameful, guilt-tripping language toward users if they stop the timer early.
Lack of task management: It offers no system to help make large tasks feel manageable or organized.
Our Opportunity: Dubble takes the avatar body-doubling mechanic, but makes it personal, removes all guilt-tripping language, and focuses on easing users into starting a task with structured step breakdowns rather than abruptly forcing a running timer
Qualitative Research & 4 Core Principles
Through qualitative coding of our co-design sessions, we created four design principles:
Encourage Motivation Without Gamification: Avoid turning productivity into a game; points create new performance anxiety and distraction.
Ease Users into Task Initiation: Eliminate the "blank canvas" freeze by pre-structuring tasks into immediate micro-steps.
Re-Engage Without Shame: Swap alarms and distracting banners that may cause stress or shame for neutral, supportive reminders.
Immersive Atmosphere Without Distraction: Create an environment that keeps the workspace grounded.
01 - Belonging through shared interests
Common interests can provide a natural starting point for connection.
02 - Social stigma
Students may hesitate to share their interests out of fear of being judged.
03 - Psychological barriers
Anxiety and uncertainty can make initiating conversations difficult.
04 - Practical barriers
Scheduling conflicts and difficulties meeting people in person can limit opportunities to connect.
Ideation & Co-Designer Critique
Each team member developed wireframe concepts across 6 directions (ranging from interactive birds to note-taking dashboards). During our co-design review:
The co-designer strongly gravitated toward the customizable avatar body-double concept.
They loved the emotional investment of building their "little guy," but emphasized that the avatar needed to actively reflect their work state rather than just sit as static decoration.
My Task: Architecting the Task Initation Flow
To solve the task initiation barrier, I built the Task Creation workflow around cognitive scaffolding:
One-Click Imports: Integrated Canvas and Google Classroom entry points to eliminate the barrier of manual data entry.
Auto-Generated Preset Steps: Selecting a task category (e.g., "Paper", "Problem Set") automatically populates editable sub-steps, transforming an overwhelming 5-hour project into immediate 15-minute milestones.
Inline Step Editing: Full user agency to modify, reorder, or delete steps before launching.
Insert information architecture here
Interactive Design of Body Double
A big id
Insert information architecture here
The Design Process
Qualitative Analysis & 4 Guiding Principles
We evaluated our high-fidelity prototype through a structured qualitative feedback protocol (Remove, Adapt, Amplify, Add):
High Emotional Resonance: The co-designer celebrated the companion mechanic:
"The little guy is pretty fire... The biggest selling point is making your little guy."
Non-Intrusive Presence: The floating reminder bubbles succeeded in redirecting focus without triggering defense mechanisms:
"Nothing immediately springs to mind [as stressful]. Maybe 5 minutes might be too short for pop-ups, but it is just a pop-up, click and done. I don't think anything is intrusive or bad."
Design Pivots Identified in Testing:
Add Task Button Placement: The co-designer hesitated when initiating tasks because the button sat isolated at the bottom-right. We moved task initiation inline with primary top-level functions so their gaze landed on it immediately.
Timer Anxiety Toggle: Watching seconds tick down on the minimized widget induced stress for our co-designer. We introduced a setting to toggle the timer off, keeping only the avatar visible.
Audio Cues for Physical Tasks: For offline work away from the monitor, they requested soft ambient audio chimes accompanying floating bubbles to gently call them back without jarring alarms
Through qualitative coding of our co-design sessions, we distilled four operational principles:
Encourage Motivation Without Gamification: Avoid turning productivity into an RPG; points create new performance anxiety and distraction.
Ease Users into Task Initiation: Eliminate the "blank canvas" freeze by pre-structuring tasks into immediate micro-steps.
Re-Engage Without Shame: Swap alarms and red banners for neutral, supportive check-ins.
Immersive Atmosphere Without Distraction: Provide calm background environments and ambient soundscapes that keep the workspace grounded.
[FEEDBACK CARD] Summary of Co-Designer Feedback across Remove / Adapt / Amplify / Add
Ideation, Concept Sketches & The Pitch
I explored initial concepts through rapid paper sketches, brainstorming ways to pair task management with ambient companion presence:
My Conceptual Pitch (Task Breakdown): I pitched the concept that the web application should be able to break tasks down in digestible steps so that users could easily know what to do next without feeling intimidated or lost. That way, they would be able to track their progress overtime on a task
Down-Selecting Ideas: Reviewing concepts with our co-designer eliminated passive journaling and note-taking concepts in favor of the customizable body-double companion and guided task setup.
[IMAGE] Concept Ideation Sketches: Early paper explorations of the mirroring avatar │
Paper Prototype to Mid-Fidelity
The Failure on Paper: My initial paper wireframe for task creation looked like every standard to-do list: a blank input field where the user had to manually type out their sub-tasks. Our co-designer pointed out that staring at a blank screen to plan a task triggered the exact same initiation dread as starting the task itself.
The Mid-Fi Pivot: I converted the task creation screen into an automated scaffolding engine. I introduced LMS imports (Canvas and Google Classroom) and category presets that automatically generate suggested sub-steps based on the task type (e.g., selecting Essay automatically suggests Research Sources → Outline → Rough Draft).
[IMAGE] Side-by-Side: Blank Paper Wireframe vs. Mid-Fi Auto-Scaffolding Screen
Mid-Fidelity Critique to High-Fidelity Refinements
During usability critiques of our mid-fidelity prototype, I identified two key points of friction and resolved them in high fidelity:
Relocating Task Creation: In my mid-fi layout, the "Add Task" button was placed in the bottom-right corner. During walkthroughs, our co-designer hesitated because their eyes naturally scanned the top navigation. I relocated task initiation inline with primary top-level functions so it was immediately visible.
Grounding the Avatar: The avatar initially floated in an empty grey void, making it feel detached. I introduced a minimalist, calm background environment to create visual immersion without introducing sensory clutter.
Editable Step Presets: I added inline Edit toggles directly beside every auto-generated step so users could adapt the generated milestones in one click before confirming.
[IMAGE] UI Evolution: Button placement relocation and background grounding
The Prototype
Outcomes of Dubble
What Dubble Solved
Bypassed the "Cold Start": In testing, automated preset steps completely eliminated our co-designer's blank-page freeze. Tasks felt approachable before the timer ever began.
Genuine Companion Attachment: The avatar provided grounding without social performance anxiety:
"The little guy is pretty fire... The biggest selling point is making your little guy."
Low-Stakes Nudges: Tactile floating bubbles redirected focus without triggering guilt:
"It is just a pop-up, click and done. I don't think anything is intrusive or bad."
[FEEDBACK CARD] Summary of Co-Designer Feedback across Remove / Adapt / Amplify / Add
Unresolved Challenges
Timer Anxiety Toggle: Our co-designer revealed that watching seconds tick down on the companion widget caused acute visual stress. We identified that the timer must have an on/off toggle so users can focus solely on the avatar.
Offline Audio Needs: For non-digital tasks (cleaning, physical reading), visual bubbles were missed. A subtle audio chime is needed to support work away from the screen.
The 1-to-1 Mirroring Scope: Due to the 9-week sprint limit, we could not build my full vision of dynamic task animations (cooking, reading, writing), keeping the avatar at a desk workspace for the prototype.
Learnings & Next Steps
Gamification doesn't work for everyone
Most productivity apps assume adding points, levels, and timers thinking it makes work engaging. I fell into this exact trap early on by projecting what I thought would be motivating. But our co-designer showed me the reality: a ticking timer didn’t motivate them—it caused them stress, and avatar leveling became another distraction. Through this project, I learned that inclusive design does not mean guessing what a group needs based research or assumptions. Engaging a co-designer exposed that traditional productivity mechanics (streaks, gamification, timers) often do more harm than good for neurodivergent minds.
The Value of Versatile User Experiences
A feature designed to help can easily backfire depending on who is using it. While a visible timer helps some people stay on track, our co-designer pointed out that watching the clock run down actually triggered intense anxiety for them. Testing reinforced that no two neurodivergent learners work the same way. Instead of deciding how someone "should" focus, good design provides flexibility—giving users the agency to toggle timers, customize prompts, and adapt the workspace to their own mental model.
UX should be neutral
Accessibility isn't just big buttons and readable text. It’s making sure the app doesn’t guilt-trip you with broken streaks or annoying alarms when you get distracted. A good app welcomes you back without judgment.
What I Would Build Next:
Expand the avatar physical interactions with the UI
Build out custom asset imports so users can draw or upload custom accessories for their avatar.
Test with a broader neurodivergent cohort to evaluate individual variance in notification timing and soundscape preferences.
Given more time, my primary next step would be implementing my original vision for dynamic 1-to-1 activity mirroring—allowing the body double's animations, desk tools, and environment to directly reflect the physical task the user is performing (writing, reading, or cleaning), expanding the feeling of true shared presence.
Designing With vs. Designing For
Through this project, I learned that inclusive design does not mean guessing what a group needs based research or assumptions. Engaging a co-designer exposed that traditional productivity mechanics (streaks, gamification, timers) often do more harm than good for neurodivergent minds.
Gamification doesn't work for everyone
Most productivity apps assume adding points, levels, and timers thinking it makes work engaging. I fell into this exact trap early on by projecting what I thought would be motivating. But our co-designer showed me the reality: a ticking timer didn’t motivate them—it caused them stress, and avatar leveling became another distraction.
The Value of Versatile User Experiences
A feature designed to help can easily backfire depending on who is using it. While a visible timer helps some people stay on track, our co-designer pointed out that watching the clock run down actually triggered intense anxiety for them. Testing reinforced that no two neurodivergent learners work the same way. Instead of deciding how someone "should" focus, good design provides flexibility—giving users the agency to toggle timers, customize prompts, and adapt the workspace to their own mental model.
UX should be neutral
Accessibility isn't just big buttons and readable text. It’s making sure the app doesn’t guilt-trip you with broken streaks or annoying alarms when you get distracted. A good app welcomes you back without judgment.
What I Would Build Next:
Expand the avatar physical interactions with the UI
Build out custom asset imports so users can draw or upload custom accessories for their avatar.
Test with a broader neurodivergent cohort to evaluate individual variance in notification timing and soundscape preferences.
Given more time, my primary next step would be implementing my original vision for dynamic 1-to-1 activity mirroring—allowing the body double's animations, desk tools, and environment to directly reflect the physical task the user is performing (writing, reading, or cleaning), expanding the feeling of true shared presence.