Every app badge cites this effect.
Most are citing the wrong one.

The Zeigarnik effect, as it's usually described, is the claim that people remember unfinished or interrupted tasks better than ones they've completed. It's one of the most frequently cited findings in UX and marketing writing, invoked to explain everything from unread badges to cliffhangers to streak counters.

It's also, by the most recent large-scale evidence, not a reliable finding at all. That's a genuinely interesting twist, because the effect people actually experience when a badge nags at them is real. It just isn't the effect most articles say it is.

The story starts with Kurt Lewin, the Gestalt psychologist supervising Bluma Zeigarnik at the University of Berlin. Lewin noticed that waiters at a local cafe could recall the details of unpaid tabs with real precision, but once a bill was settled, the same order vanished from memory almost immediately.

Zeigarnik turned that observation into a formal experiment, published in 1927 in the journal Psychologische Forschung. Participants worked through a series of separate tasks, puzzles, assembling a box, and similar small assignments. For roughly half the tasks, the experimenter subtly interrupted the participant partway through; the rest were completed without interruption. Her finding: participants recalled details of the interrupted tasks about 90% better than the ones they'd finished. She explained it through Lewin's field theory, arguing that starting a task creates a kind of psychological tension that stays active until the task is resolved, keeping its details more accessible in the meantime.

The trouble started early. By 1951, researcher Carl Hovland was already noting that few investigators could unequivocally reproduce Zeigarnik's original findings, and that results seemed to swing dramatically depending on the personality of whoever was being tested.

A 2025 meta-analysis put a much finer point on it. Pooling the available research on interruption, recall, and resumption, the researchers found no memory advantage for unfinished tasks at all. Their conclusion wasn't that Zeigarnik fabricated anything, but that the effect, if it exists, depends heavily on situational details and individual differences that were more common in the original lab conditions than in typical modern replications, which makes it, at best, inconsistent, and at worst, simply not a stable phenomenon.

Here's where it gets genuinely interesting: the same 2025 meta-analysis that found no memory advantage did confirm a separate, related finding, a general tendency to resume interrupted tasks. That specific claim belongs to Maria Ovsiankina, a colleague of Zeigarnik's in Lewin's Berlin lab, who studied task interruption from a behavioral angle rather than a memory one.

The distinction matters more than it sounds like it should. The Zeigarnik effect is a claim about memory: you remember unfinished things better. The Ovsiankina effect is a claim about behavior: an unfinished task creates what researchers call a "quasi-need," an active pull to go back and finish it, independent of whether you remember the details well. The meta-analysis found the memory claim doesn't hold up. The behavioral pull to resume does.

UX and marketing writing has settled almost entirely on "Zeigarnik effect" as the explanation for unread counts, app icon badges, Netflix auto-playing the next episode, and Duolingo-style streaks. It's a catchy, well-known name, and it fits the vibe of "unfinished things nag at you."

Most of that writing is citing the wrong psychologist. A red badge on an app icon isn't primarily exploiting a memory advantage, you don't need to remember details about an unread message for the badge to bother you. It's exploiting the resumption urge, the itch to close an open loop, which is Ovsiankina's finding, not Zeigarnik's. The industry picked the more famous name and ran with it.

None of this is an argument that unread badges are harmless because the popular explanation cited the wrong study. Whichever psychologist's name actually belongs on it, the practical finding that survived a 2025 meta-analysis is real: an unresolved task creates a genuine pull to go finish it. That pull is what a badge, a streak, or an unfinished feed is built to trigger, regardless of which 1920s Berlin lab first described the mechanism underneath it.

The resumption urge needs an open loop to attach itself to. An unread badge, an unfinished feed, a pending notification, all of it sitting visibly on your home screen, actively inviting resolution. Trying to feel the pull and just not act on it, over and over, all day, is a harder fight than it needs to be.

Fella keeps apps blocked by default, so those open loops aren't camped out on your home screen all day generating that pull in the first place. The one 5-minute unlock a day becomes the single, deliberate window where you can go close out whatever's actually waiting, instead of an all-day standing invitation for the urge to fire off every time you glance at your phone.