⊞ Desktop C:\Projects\FlowState\case-study.exe

UX / UI design · MSc Design Engineering, Imperial College London

FlowState

A phone app that pauses whichever apps you choose until you've finished the task you're avoiding. You earn your break with proof of work, not by waiting out a timer.

Role
Solo designer: research, UX, UI, testing
Module
User Interfaces, Dyson School of Design Engineering
Date
March 2026
Tools
Figma, paper prototyping, think-aloud testing
Three FlowState screens: task breakdown, the Instagram-is-paused intervention, and the break-earned celebration
~80%of university students regularly procrastinate
4semi-structured interviews with grad students
6think-aloud usability tests on the hi-fi prototype
60–90sfrom the "paused" screen to an earned break
📄 01_problem.txt
01

Blocked time isn't productive time

"Just five more minutes" turns into hours. Social apps are built for dopamine-driven scrolling, and for students the cost is missed deadlines, anxiety, guilt and bad sleep.

The existing tools come in two separate camps:

  • Phone-reduction apps (Jomo, Opal, Forest) block apps for a set time. The timer runs out whether or not you did any work.
  • Task tools (Notion, Pomodoro timers) organise the work but do nothing when you reach for your phone, and they rely on the honour system to tick tasks off.

The gap: no existing tool checks that a task is actually done before giving phone access back. They all use time-based restrictions. None use completion-based accountability.

Comparison table of Jomo, Opal, Forest, Notion and Pomodoro across intervention, key limitation, sensing and verification method
Competitive analysis: none of the five verify task completion
🔍 02_research.log
02

What students told me

I ran semi-structured interviews with four participants (two master's students, one PhD student and one graduate) about how they procrastinate, when they reach for their phone and what they've already tried. Six themes came out of it:

THEME 1

Starting is the hard part

Long to-do lists trigger avoidance. People need help to start a task, and organising it isn't enough.

THEME 2

Phones as a pacifier

People pick up the phone automatically at friction points. Knowing they do it doesn't stop them.

THEME 3

Timers don't verify

Everyone who'd tried Forest or Pomodoro said the same thing: the time passes whether you work or not.

THEME 4

Data breeds guilt

Screen-time stats make people feel bad but don't tell them what to do next.

THEME 5

Support in the moment

They wanted an intervention at the moment they give in, and they wanted it to feel supportive rather than controlling.

THEME 6

Earn it by finishing

Unlocking apps by completing tasks was seen as motivating, as long as it suggests a next step.

"My problem mainly is starting. It takes me ages to get started, but once I do, I'm good to go. But by then, usually most of the day has gone."

— Participant 3

"As soon as I reach a friction point… it's like a pacifying almost. I pick up my phone instead of doing what I want to do."

— Participant 2
🎯 03_design_aims.md
03

Design aims

  • Help people start, not just plan. Break big tasks into small, timed steps.
  • Intervene at the moment of procrastination, so reaching for the phone becomes a choice.
  • Tie phone access to completion, not elapsed time, verified with photo proof or a friend's code.
  • Show progress and next steps instead of guilt-inducing statistics.

Usability goals

Effective: tasks actually get done Supportive: no guilt, no punishment Memorable: works during autopilot moments Efficient: the productive path is the easy path
✏️ 04_process/
04

From flows to paper to pixels

Information architecture

The whole system is built around one moment: when someone tries to open a blocked app. Instead of a dead-end "blocked" screen, the flow runs forward through intervention → action → verification → reward, with an escape hatch back at every step.

Paper prototyping

Sticky-note prototypes let me test the flow before worrying about any visual design. Two changes came straight out of these sessions:

  • Three buttons became two. "Task complete" and the override key were merged, leaving I've completed my task and Break it down for me. One tester felt it was "a lot for the one page".
  • The current task got its own card, with a bordered, colour-accented card at the top, so the user's commitment is the first thing they see.

Lo-fi wireframes

Greyscale wireframes pinned down the hierarchy, spacing and structure before any colour went in.

Lo-fi intervention screen
Intervention
Lo-fi task breakdown
Breakdown
Lo-fi focus timer
Focus timer
Lo-fi verification choice
Verify
Lo-fi upload screen
Upload
Lo-fi success screen
Success
Table of wireframe refinements: radio buttons, icons with labels, large centred timer, lock badges
Wireframe refinements: initial approach vs. final decision
🎨 05_visual_identity.fig
05

Colour with a job to do

Each colour has one meaning across the whole app, which makes the flow easy to read at a glance. The palette also avoids relying on red/green, so it still works for colour-blind users.

Coral = actionEnergising and optimistic. Used for the intervention screen, headers and primary CTAs.
Purple = focusLinked to calm and creativity. Used for task breakdown and selection.
Teal = successGrowth and achievement. Used for verification and the break-earned state.

Type & components: Inter, on an iOS HIG-style scale (28 / 17 / 13 px). Buttons use a 16 px radius with gradient fills and cards use 18 px. All interactive elements meet WCAG 2.1 AA contrast.

📱 06_final_design.fig
06

The happy path

Users choose which apps FlowState pauses. In this walkthrough it's Instagram: Marcus has a report due and opens it out of habit. FlowState steps in, helps him pick one small piece of the report, checks he's done it, and gives him 15 minutes of guilt-free scrolling.

Phone home screen with an Instagram notification
1 · TemptationA notification or an app tap both trigger it
Intervention screen: Instagram is paused, progress ring and two choices
2 · Intervention"Instagram is paused", not "blocked"
Task breakdown screen with subtasks and time estimates
3 · Break it downSmall subtasks with time estimates
Focus timer with blocked app icons
4 · FocusThe timer is information, not a lock
Verification method choice
5 · VerifyUpload proof or a friend's social key
Upload proof screen
6 · ProofPhoto of the work, checked by AI
Break earned celebration with 15 minute timer
7 · Reward15:00 of full access, or keep going
▶ prototype.fig

Click through it yourself

This is the high-fidelity Figma prototype of the happy path. Tap the Instagram notification to start.

🧪 07_testing.log
07

Testing & iteration

Six participants worked through four core tasks on the hi-fi prototype while thinking aloud. The core idea held up:

"Oh, that's so nice. Now I can know what I need to do next. That makes it far easier to tackle the task."

— on "Break it down"

"I felt good not opening Instagram because it reminded me not to."

— on the intervention screen

Several participants said they'd genuinely use it. One pushed back ("locking out is good enough… I'm just being lazy"), which suggests the breakdown feature helps most with starting tasks, less with simple restriction.

Participant testing the hi-fi prototype on a phone next to a laptop
Think-aloud testing on the hi-fi prototype

What changed

Issue observedDesign changeWhy
Testers tapped the notification, not the app iconBoth entry points trigger the interventionUsers reach apps through many routes
Testers tapped the upload preview, not the buttonThe whole preview area is now tappableMatch the mental model; bigger target (Fitts)
"Photo" and "File upload" felt identicalMerged into one Upload proof optionLess cognitive load (Hick's law)
💭 08_reflection.txt
08

Where it could go

PRIORITY 1

Generative AI

Breakdowns that learn from past completion rates, a way to type "I'm stuck on the methodology" and get targeted help, and checking whether an upload shows real progress.

PRIORITY 2

Wearables & smart home

A watch complication and gentle haptics so you don't have to look at your phone to avoid your phone. Lights could shift colour during focus sessions.

LATER

AR / VR

Project the current task onto your desk. Any VR focus room has to lead back to the real work, not become another way to escape it.

Looking back

I'm proudest that the research backed up the core idea: participants recognised the problem in their own lives and asked to use the product. Next time I'd widen the participant pool beyond postgrads, since procrastination also covers chores and working life, and I'd design the onboarding flow, calendar integration and adjustable strictness levels.