Pair something people want to do with something they know they should do — and the should becomes something they actually look forward to. The temptation makes the obligation tolerable. The obligation gives the temptation a legitimate home. Used deliberately, this is one of the most effective mechanisms for driving behaviour that users resist when asked for it directly.
In 2013, behavioural economist Katherine Milkman ran an experiment at a university gym. She gave participants access to audiobooks they loved — but only while working out. The audiobooks could not be listened to anywhere else. Gym attendance increased by 51% among participants who received the most tempting audiobooks.
Milkman called this temptation bundling: the deliberate pairing of an instantly gratifying “want” activity with a less immediately rewarding “should” activity. The bundle works because humans are present-biased — immediate rewards are weighted far more heavily than future ones.
The principle translates directly into product design. Every product contains both “want” moments (discovery, achievement, social recognition, beautiful interfaces, surprise) and “should” moments (completing a profile, connecting integrations, reviewing data). Most products sequence these badly — front-loading the should and back-loading the want. Temptation bundling inverts the relationship.
“The key insight is to only indulge in a tempting activity while doing the thing you should be doing.”
Onboarding is the place where most products most consistently fail at temptation bundling. The typical onboarding flow is a sequence of should-tasks — connect your calendar, invite your team, set your preferences — with the product's actual value deferred until after they are complete. The user experiences obligation before they have experienced anything worth the obligation.
The want in onboarding is seeing the product come alive around the user's own data. The should is connecting that data. The bundle: deliver the visual payoff at the exact moment each connection is made.
Three chores, no reward. Dashboard is blank until all integrations are connected. Continue is disabled. Pure obligation.
Calendar is already connected and showing real data. Remaining integrations are visible as ghost panels on the live dashboard. The want is immediate.
Habit formation is where temptation bundling has its most visible product applications. Duolingo, Peloton, Headspace, and Strava all embed the same mechanism: a repetitive action that is not intrinsically rewarding (practising vocabulary, exercising, meditating, logging a run) is bundled with something that is (streak maintenance, character reactions, leaderboard position, social sharing).
The key insight is structural: the fun part only exists inside the effortful part. You cannot maintain a Duolingo streak without doing a lesson. You cannot see your Peloton leaderboard position without completing a ride. The want is gated by the should — not as a reward for completing it, but as something that happens simultaneously with it.
Language practice with no immediate emotional reward. The benefit is abstract and distant.
Same practice, but bundled with streak counter, character celebration, and XP accumulation. The effort and reward are simultaneous.
Temptation bundling is not gamification. It is a structural decision about when and how the rewarding parts of a product experience are delivered relative to the effortful parts. The question for every “should” moment in your product is: what is the “want” that can happen simultaneously with it — not after it?
Milkman, K. L., Minson, J. A., & Volpp, K. G. (2014). Holding the hunger games hostage at the gym. Management Science, 60(2), 283-299. Milkman, K. (2021). How to Change. Portfolio/Penguin.