Temptation bundling
Pair something you should do with something you want, and the should-do happens more.
Temptation bundling restricts an indulgence so it is only available while performing a beneficial but unappealing activity. The indulgence supplies the immediate pull that the beneficial activity lacks, and the restriction keeps the pull attached. It depends on the restriction holding, which is where it usually breaks.
How it shows up in software
Bundling shows up when a product attaches an enjoyable surface to a chore: music that plays only during a workout, a podcast that resumes only on a commute, a game layer sitting on top of a practice drill. In software the restriction is technically easy and socially hard, because users dislike being told where they can consume their own content.
Using it well
- Let the user choose the indulgence, since the pull only exists if they actually want the thing.
- Make the restriction light and honest: reserve a curated set rather than locking content the user already owns.
- Tie the bundle to a context rather than a threshold, so it feels like an occasion rather than a gate.
- Expect decay, and check whether the bundle still moves behaviour after a month before you build more on it.
Where it turns manipulative
- Withholding content the user has already paid for until they perform a behaviour you want is a hostage mechanic wearing a research citation.
- Bundling an addictive surface onto a health behaviour can leave the user with a new compulsion and no lasting change in the behaviour you targeted.
- Applying it to a behaviour the user did not ask to change, such as bundling entertainment with ad viewing, is advertising, not motivation design.
Where you have seen it
Peloton
Artist-specific rides put music people want behind a workout they are ambivalent about.
Zombies, Run!
Story audio advances only while the user is running, tying narrative to activity.
What the research says
- Milkman, Minson and Volpp, 2014Mixed evidence
Gym-goers given audiobooks available only at the gym attended more often than controls, with the largest effect for those whose access was fully restricted.
The effect decayed over the study and largely vanished after a Thanksgiving break, which the authors report. The persistence claim is weaker than the headline.
- Kirgios, Mandal, Talaifar and colleagues, 2020Mixed evidence
A large field test of bundling audio entertainment with gym visits found effects smaller than the original study and concentrated in the period while support continued.
The usual pattern: an appealing lab or pilot result shrinks at scale.
Grades are a judgement about the evidence, not about how useful the idea is. Plenty of contested effects are still worth knowing, as long as you do not cite them as settled.