Unlock it, or don't lose it — the two ways to word the same nudge.

Take the plainest nudge you have — the one that asks a half-set-up user to finish. You can word it two ways that are, on paper, the same request. “Finish setting up to unlock your dashboard.” Or: “You’ll lose your progress if you don’t finish setting up.” Same action, same user, same screen. The only thing that changed is whether you pointed at the good thing that comes from acting or the bad thing that comes from not acting.

Psychologists call that a goal frame, and it’s worth being precise, because “framing” gets used to mean three different things. Irwin Levin and colleagues untangled them in 1998. There’s risky-choice framing (the classic surgery-survival-vs-mortality gambles). There’s attribute framing (“90% lean” vs “10% fat”). And there’s goal framing — the one you’re actually using in a nudge — where you frame the same behavior by the benefit of doing it or the cost of skipping it. That’s the lever in play every time you write onboarding copy, a re-engagement email, or a “you left something behind” banner.

Now, the honest part, because this is where growth folklore gets ahead of the evidence. There is a real asymmetry in how brains weight the two, and it leans toward loss. “Bad is stronger than good” is close to a law in psychology — Roy Baumeister’s team catalogued it across memory, emotion, learning, and impressions; a threatened loss grabs attention harder and faster than an equivalent gain. That’s why the “you’ll lose your draft” line makes more people look up. But — and this is the part the folklore drops — the behavioral edge from goal framing is modest and context-dependent, not the 2× some playbooks imply. The cleanest evidence comes from health messaging, where Gallagher and Updegraff’s 2012 meta-analysis found loss frames did slightly better for detection behaviors (get the scary screening) and gain frames slightly better for prevention behaviors (keep up the good habit) — and even those differences were small. Treat any single A/B win here as noisy until it repeats.

So the practical rule isn’t “loss frames convert better.” It’s match the frame to the behavior, and spend the loss frame like it’s expensive — because it is.

Reach for loss framing when the thing at risk is real, concrete, and the user’s — an unsaved draft, an expiring trial they actually value, a one-time window that genuinely closes. There the loss frame isn’t manipulation; it’s information they’d want. Reach for gain framing for the ongoing, voluntary, aspirational behaviors — the habit you want them to keep, the feature you want them to explore, the reason to come back tomorrow. Those wear better when they pull toward something good than when they’re policed by a threat.

⁂

And watch the failure mode, because loss framing has a nasty back edge. Lean on it for things that aren’t really losses — “don’t miss out,” “your account may be at risk,” manufactured countdowns — and you win the glance but spend trust every time. A brain that keeps getting told it’s about to lose something, and keeps not losing it, recalibrates. The frame that grabbed attention in week one reads as noise by week four, and the product that cried loss gets muted with everything else.

The two wordings are never neutral, and they’re never quite equal. But the difference is a small, honest lever, not a growth cheat code. Frame the gain for what you want them to keep doing; frame the loss only when something true is genuinely on the line — and let the modesty of the effect keep you from turning every nudge into an alarm.