Zeigarnik Effect
People remember uncompleted or interrupted tasks better than completed ones — creating a cognitive pull to finish.
Plain English
The Zeigarnik Effect is the psychological tension of an unfinished task. When we start something and do not finish it, it occupies mental real estate — an open loop in working memory that keeps nagging for closure. This is why progress bars, checklists, and streaks are so effective at driving re-engagement: they make incompleteness visible and therefore cognitively uncomfortable. It is also why showing a user 'Profile 60% complete' pulls them back to the settings page. The dark side: notification badges, endless feeds, and streak mechanics exploit this same mechanism to drive compulsive behaviour.
Research background
Discovered by Bluma Zeigarnik (1927), a student of Kurt Lewin, who found that waiters remembered unpaid orders better than paid ones — an open transaction creates a cognitive 'task set' that is maintained in working memory until closed. In UI: (1) Progress indicators — showing partial completion creates a motivation to complete. (2) Onboarding checklists — each unchecked item is a Zeigarnik tension. (3) Streaks — breaking a streak feels like unfinished business. (4) Draft saving — 'You have an unsent draft' re-engages users. Ethical boundary: progress indicators should reflect genuine progress, not manufactured anxiety. Streak mechanics should have forgiveness (streak freeze) to avoid distress.
UI examples
✓ Applying the principle
LinkedIn's profile strength meter ('You're at Intermediate! Add a skill to reach All-Star') — a permanently incomplete progress indicator that drives repeated return visits to complete profile sections.
✗ Violating it
A SaaS onboarding checklist where 8 of 10 items are optional but shown as incomplete — creates Zeigarnik anxiety without delivering value from completion, training users to ignore progress indicators.
AI Prompt
Paste into Claude, Cursor, Bolt, or v0.
Use the Zeigarnik Effect ethically: add a profile completion bar (real progress only), an onboarding checklist of 5 high-value actions, and a 'continue where you left off' re-engagement prompt. Avoid manufacturing false incompleteness — only show open loops that close into genuine user value.