The Audiobook Trick That Got People to the Gym
A Wharton professor's personal habit of only listening to guilty pleasure audiobooks at the gym became a formally tested technique called temptation bundling. The research found it genuinely worked, along with an honest limitation worth knowing before relying on it.
A few years into her career, Wharton professor Katy Milkman noticed she had a reliable weakness, addictive, not particularly literary audiobooks, the kind she would happily listen to for hours if given the chance, and a much less reliable relationship with the gym. She started allowing herself to listen to those guilty pleasure audiobooks only while exercising, nowhere else. Within a short time, she was going to the gym five days a week. Years later, she turned her own accidental discovery into a formal field experiment and gave it a name, temptation bundling.
The study, published with Julia Minson and Kevin Volpp, took the idea out of one person's personal life and tested it properly. Participants were given iPods loaded with genuinely tempting audio novels, then randomly assigned to one of three conditions. One group could only access the audiobooks at the gym, their access physically restricted through the device itself. A second group was simply encouraged to try limiting the audiobooks to gym time, on the honor system, without any actual restriction in place. A third group served as an ordinary control, with unrestricted access throughout.
The results were clear in direction if not permanent in size. The fully restricted group visited the gym fifty one percent more often than the control group. The honor system group, asked to self restrict without any actual barrier, still visited twenty nine percent more often than control, showing the bundling idea itself carried some real weight even without hard enforcement behind it. Both effects faded somewhat as the study went on, dropping off noticeably after the Thanksgiving holiday disrupted routines generally, a reminder that even a genuinely effective technique is not immune to the ordinary forces that derail habits around any major disruption to someone's routine.
One detail from the study is arguably more interesting than the headline result itself. After the experiment ended, sixty one percent of participants chose to pay their own money for continued restricted access to their tempting audiobooks, want only at the gym, rather than simply take the freely available unrestricted version back for nothing.
One detail from the study is arguably more interesting than the headline result itself.
Milkman and her collaborators have since tested variations on the same underlying idea in other settings, including workplace wellness programs that paired tempting content with healthy behaviors more broadly, generally finding the same basic pattern, a real and meaningful boost in the target behavior that is genuine but not permanent, and that works better when the restriction has some actual teeth to it rather than relying purely on a person's own willingness to self impose the limit. The consistency of that pattern across different studies is part of why temptation bundling gets taken seriously as a specific technique in behavioral economics rather than dismissed as one clever anecdote that happened to work for one professor.
That is a strange thing to pay for on its face, less access to something you enjoy, and it points to a genuine, widely shared frustration with plain willpower as a strategy. Given the option, a majority of people preferred to pay for a structure that removed a choice rather than trust themselves to keep making the harder choice unprompted, every single time, for free.
The underlying logic generalizes well beyond audiobooks and gyms. Temptation bundling works by taking something a person already wants to do regardless of any self improvement goal, and locking it specifically to a should behavior that tends to lose out when it has to compete against willpower alone. A favorite podcast reserved only for a work commute you have been avoiding walking instead of driving. A specific streaming show allowed only while doing the household task that always gets put off until it becomes urgent. The want does not have to be virtuous or productive on its own terms. Its only job is being strong enough, and specific enough, to reliably show up and pull the should behavior along with it.
The honest caveat the research itself surfaces is that this is a genuine tool, not a permanent fix that stops working once installed and forgotten about. Bundles wear out, holidays and disruptions break the pattern, and a want that stops feeling tempting stops doing any work at all. The practical takeaway is less about finding the one perfect bundle and more about treating the pairing itself as something to notice when it starts fraying, and to rebuild deliberately with a fresh want once the old one has clearly lost its pull.