ScreenSage
A gamified approach to reducing phone screen time through social accountability and friendly competition.
Passive trackers fail because people don't respond to data about themselves. They respond to how they compare to people they care about.
"I don't care about the graph, I care if I beat my friend."
ScreenSage · user interview
Awareness isn't enough.
Passive well-being tools show you a graph of your failure, then offer no reason to change.
Passive tools · observed failure modes
Turning a personal habit into a shared challenge.
"Limit yourself."
→"Beat your friends."
Weekly Leagues
3 to 5 friends. Resets every Monday.
Cumulative Scoring
Total screen time ranks the squad. Lowest wins, live.
Discovery
"Don't ask them what they want."
Jobs-to-be-Done interviews: No pitching, just their last week with the phone.
Question asked
"Tell me about the last time you felt you spent too much time on your phone."
Insight Uncovered
Doom-scrolling spiked while waiting or feeling lonely. The missing element was social.
What the research overturned
Global leaderboards
→Squads of people you know. Your roommate matters, strangers don't.
Infinite streaks
→Weekly resets. A bad day shouldn't end the game.
Locking apps
→Positive competition. Locks drove uninstalls, play drove engagement.
ScreenSage · discovery interviews
What We Built
Scoped to the core loop to fit a tight sprint timeline.
1. Onboarding
Screen-time permissions, then a profile.
2. Squad Creation
Invite-only. You compete with people you know.
3. The Leaderboard
Live rank by screen time. Lowest wins.
Learnings & Reflection
Talk Less. Observe More.
Real behavior surfaced more than stated preferences ever did.
Psychology over Features
Behavior change is a motivation problem, not a feature problem.
Validate Early
Proving the core mechanism beat building a full feature suite.
Iterate on Friction
Cutting onboarding to one step, "Join a Squad," lifted conversion roughly 40%.
ScreenSage · prototype tests
The right insight unlocks the right product.
Limits, notifications, and graphs had all been tried. One interview question showed the problem was motivation, not information.