Your points system might be lowering engagement

In an early-1970s study, researchers found a group of preschoolers who already loved to draw — kids who reached for the markers on their own, for no reason but that drawing was fun. They split them. One group was told they’d get a fancy “Good Player” certificate for drawing; they drew, and got it. Another group just drew, no reward. Weeks later, back in free play, the kids who had been rewarded drew noticeably less than the ones who hadn’t. The certificate hadn’t added motivation. It had quietly eaten some.

This is the overjustification effect, and it’s worth sitting with because it runs against the deepest assumption in product design: that motivation adds up. More reasons to do a thing means more doing, right? Bolt points onto the behavior and you’ve stacked an external reason on top of whatever internal reason was already there. Except the mind doesn’t stack reasons. It looks at its own behavior and infers why — and when a big, obvious external reward shows up, the inference shifts. I used to do this because I liked it. Now I’m doing it for the reward. Once that reframe happens, the original reason atrophies from disuse, and the day the reward disappears — or just stops feeling novel — there’s nothing underneath holding the behavior up.

You can see the failure this sets up. A habit app that people genuinely used because journaling helped them adds an aggressive streak counter. Usage climbs — for a while. But now the reason to open it has quietly migrated from “this helps me think” to “I can’t break my streak,” which is a more brittle, more anxious reason. Then one busy day the streak snaps, and instead of shrugging and journaling anyway, they quit entirely, because the thing they were actually doing it for — the number — is gone, and the older reason withered while the streak was doing the work. The extrinsic scaffold collapsed and took the intrinsic motive down with it.

The important nuance, and the thing that keeps this from being “all rewards are poison,” is that overjustification is conditional. It bites hardest when two things are true: the behavior was already intrinsically motivated, and the reward is controlling — expected, contingent, “do this to get that.” It barely applies, and can even help, when the task had no intrinsic pull to begin with (there’s nothing to crowd out — extrinsic motivation is all you’ve got, so use it) or when the reward is unexpected and informational rather than controlling — a surprise thank-you, feedback that tells someone they’re getting good at the thing. A reward that says you’re becoming competent feeds the internal reason. A reward that says here’s your payment replaces it.

So the design rule isn’t “never reward.” It’s know which motivation you’re feeding, and reward the right layer. Prefer rewards that signal progress and mastery over rewards that feel like currency — “you’ve learned 200 words” reinforces why someone came; “earn 50 gems” slowly overwrites it. Make the core loop worth doing for its own sake rather than propping a hollow core up with points, because points on top of something genuinely fun is where the crowding-out does its damage. And keep an eye on the withdrawal cliff: if this week’s engagement leans on a reward, ask honestly what happens the week you remove it, or the month the novelty dies.

The obvious objection is that gamification plainly works — Duolingo’s streaks, the fitness rings people chase to close. That’s fair, and here’s the honest version: extrinsic rewards absolutely can drive behavior, especially for tasks with weak intrinsic pull, and the best-designed ones bridge a person across the dull early stretch of a hard new habit until intrinsic motivation can take over. The danger was never rewards as such. It’s rewarding what people already loved, in a controlling way, and then reading the short-term usage bump as durable motivation when it might be the opposite. Which points at the only metric that actually answers the question: not “did adding points raise this week’s numbers,” but “if we took the points away, would they still be here?”

I’ll be straight that the size of this effect is debated — some reviews find tangible, expected rewards reliably undermine intrinsic motivation, others argue the conditions are narrower than the headline suggests. Treat it not as “rewards are dangerous” but as “motivation isn’t additive, so find out which kind you’re touching.” Which means before you add the points, do the unglamorous thing and find out why people currently show up. If the honest answer is because they want to, you don’t have a motivation problem to solve. You have one to protect.