The badge that can never reach zero

A waiter can hold a dozen unpaid orders in his head perfectly and forget every one the moment the bill is settled. That observation, nearly a century old now, is the usual origin story for the Zeigarnik effect: unfinished tasks sit in memory with a faint tension attached, tugging at attention, and the tug releases the instant the task is done. An open loop is heavier than a closed one.

Builders already use this, whether or not they know its name. A progress bar is an open loop you can see. A checklist with one item ticked is an open loop with a foothold. “Resume where you left off” reopens a loop on purpose. Used well, this is genuinely kind: the tension is real, the finish line is real, and closing the loop hands the user something they actually wanted. I’ve written before about why a non-empty checklist and a visible next step accelerate completion. That version of the mechanism is a gift.

This piece is about the same mechanism pointed somewhere darker.

Because here’s the thing the effect doesn’t require: it does not require the loop to be closeable. The tension comes from the loop being open, not from there being a real reward at the end. Which means you can manufacture the pull without ever offering the release. A red badge that reappears the moment it’s cleared. A “profile 87% complete” ring that’s engineered never to reach 100. A notifications count that regenerates on a timer. An achievement track with no last rung. Each of these borrows the exact same attentional tug as a helpful progress bar — and never pays it back.

The reason this is worth naming as its own trap is that it doesn’t feel like a dark pattern from the inside. Nobody sits in a planning meeting and says “let’s rent out the user’s low-grade anxiety.” They say “engagement is up when we show the badge,” which is true, because a tug you never release keeps people checking. The metric rewards you for the open loop and is blind to the cost, which lands on the user as a background hum of unfinished business they can’t ever clear.

So here’s the test I’d put on any progress cue, badge, or count before shipping it: can the loop actually be closed, and does closing it complete something the user came here to do? A progress ring that reaches 100% and then stays at 100% is a to-do list. A ring that resets, or that was designed never to fill, is an itch you installed. The first respects the mechanism; the second exploits it.

This is distinct from a couple of neighbors worth keeping straight. It isn’t the urgency trap — a loud “act now” cue that hijacks attention by manufacturing time pressure. Zeigarnik tension isn’t about urgency, it’s about incompleteness; the badge doesn’t say hurry, it says unfinished. And it’s not quite the goal-gradient pull toward a visible finish line either — the danger here is precisely a finish line that was never real. Same family of open-loop psychology, three different failure modes.

One honesty note on the science: the original waiter finding is old, and its replications have been uneven — the effect is real but moderated by how much the person cares about finishing, and it’s not the ironclad law pop-psychology makes it. That actually strengthens the design point rather than weakening it. If the tug only shows up when the user is invested in completing, then a loop you keep open with no payoff isn’t just manipulative, it’s manipulative and unreliable — you’re taxing the exact users who trusted your finish line was real, and training everyone else to ignore your badges.