When Walter Isaacson first met Steve Jobs to discuss a panel, Jobs declined the panel and proposed a walk instead. “I didn’t yet know that taking a long walk was his preferred way to have a serious conversation,” Isaacson later wrote in his biography Steve Jobs, as recounted by CNBC. The walk ended with Jobs asking Isaacson to write the book. It is a good story, and like most good stories about productivity habits, it is usually told without the part that matters: walking is not a general-purpose upgrade to your brain. It helps a specific kind of thinking and does almost nothing for another. Founders who know the difference get something out of the practice. Founders who treat it as a cure-all are using a screwdriver as a hammer.
The claim that “I walk to think” is one of the few productivity assertions with a clean experimental backbone. Most of the genre is anecdote dressed as science. This one is closer to science dressed as anecdote.
The study everyone cites and almost nobody reads
The source is a 2014 paper by Marily Oppezzo and Daniel Schwartz of Stanford, published in the Journal of Experimental Psychology: Learning, Memory, and Cognition, volume 40, number 4, pages 1142 to 1152. The title is “Give Your Ideas Some Legs: The Positive Effect of Walking on Creative Thinking.” Per the abstract recorded by ERIC, it ran four experiments, and the design is more careful than the headlines suggest.
The researchers separated two kinds of thinking that most people lump together. The first is divergent thinking, measured with Guilford’s alternate uses test, where you generate as many uses as you can for a common object. A brick. A button. The second is convergent thinking, measured with the compound remote associates test, where there is one correct answer: find the single word that connects three given words. Divergent thinking is brainstorming. Convergent thinking is solving.
The split in the results is the whole point. According to the ERIC abstract, walking increased creativity on the alternate uses test for 81 percent of participants, but it improved convergent scores for only 23 percent. Walking opened the flow of ideas. It did not help find the right answer. CNBC, summarizing the same work, reported that researchers asked 176 college students to complete tasks while sitting and then again while walking, and found participants were “overwhelmingly” more creative when walking. The popular shorthand became “walking boosts creative thinking by an average of 60 percent,” a figure CNBC and others repeat. The cleaner number to carry around is the contrast: 81 versus 23. The same activity that helps you list options does not help you pick among them.
Two further findings are worth keeping. The ERIC abstract notes that in Experiment 2, participants who sat down after walking still showed a residual creative boost. You do not need to be moving at the moment the idea is needed. A walk beforehand charges the effect for a while. And Experiment 3 generalized the result to outdoor walking, while Experiment 4 found that walking outside produced the most novel and highest-quality analogies, with the effects of being outdoors and the act of walking separable from each other. Movement did the work. The scenery was a bonus, not the mechanism.
Why the founder accounts are unusually consistent
Most habits attributed to famous founders fall apart on inspection, because the founders did a dozen things and survivorship bias picks the photogenic one. The walking habit holds up better, partly because it shows up across people who share almost nothing else.
Jobs took serious meetings on foot, especially the important ones, and per CNBC was often seen walking the Apple campus with chief designer Jony Ive while working through designs. Jack Dorsey, the Twitter and Square cofounder, told Fortune, as quoted in a widely circulated write-up, “If I’m with a friend we have our best conversations while walking,” and reportedly walked roughly five miles to his office. Per the same account, Mark Zuckerberg and former LinkedIn CEO Jeff Weiner have used walking meetings as a standard practice. The habit is not bound to one temperament or one industry.
The creative-work version is even more on-the-nose. Aaron Sorkin, the screenwriter behind The West Wing, The Social Network, and A Few Good Men, does not write sitting at a desk in the usual sense. As described by Lound, he walks for hours, speaking dialogue out loud and playing every character, letting scenes form through movement and voice. That is divergent thinking in its purest form: generating, riffing, finding the line that does not exist yet. It is exactly the task the Stanford data says walking helps, and Sorkin arrived at the method without the study.
Then there is the older lineage, which is more than decoration. Nilofer Merchant, in her TED talk “Got a Meeting? Take a Walk,” built the case from the other end, starting with sitting rather than walking. Per the transcript, she cites figures that people sit 9.3 hours a day, more than the 7.7 hours they sleep, and argues that “sitting has become the smoking of our generation.” She converted to walking meetings after a contact could only meet while walking her dogs, and now does 20 to 30 miles a week of them. The line she keeps is practical: getting out of the box leads to out-of-the-box thinking, and walking let her hold two problems “in opposition when they’re really not that way.” That is the divergent mode again, described by someone who measured it in her own calendar rather than in a lab.
What walking does not do
The honest version of this advice names what the practice cannot deliver, and the study is unusually direct about it. Walking did not help convergent thinking. If the task in front of you has a single correct answer, sitting was the better posture for 77 percent of participants in the ERIC data. Closing a financial model. Reconciling two numbers that should match. Reading a contract clause and deciding whether it binds you. Debugging a specific failure with a specific cause. These are convergent problems, and a walk will not surface the answer. It may even cost you, because the same loosening that helps you generate options makes it harder to lock onto the one right thread.
This is the part most “walk to think” content drops, and dropping it is what turns a real finding into a productivity myth. A walking meeting is a brainstorming format. It is not a decision format. The two get conflated constantly, and the cost is meetings that end with a pleasant sense of momentum and no actual choice made. You can generate ten directions on a walk. Committing to one of them is work better done sitting down, with the options written where everyone can see them.
A walking meeting is a brainstorming format. It is not a decision format.
There is also a quieter limit. The residual boost the ERIC abstract describes is short. It fades. A walk is not a deposit you can draw against all week. It is a state you enter and then leave, useful in the window around the activity, which is why the practical move is to walk close to when you need the ideas, not the day before.
A structure that matches the evidence
The applied version is almost embarrassingly simple, which is a point in its favor. Match the format to the kind of thinking the meeting needs.
Use a walk for the generative half. Naming a new product. Listing every reason a deal might fall through. Working out the shape of a hard conversation before having it. Mapping the possible causes of a problem before you have narrowed them. These are the alternate-uses tasks of running a company, and they are where the 81 percent lived in the Stanford data.
Keep the convergent half at a desk. Once the options exist, sit down to choose. Put them in front of people. The decision benefits from the focus that walking deliberately relaxes. The two-phase split is not a compromise between the data and the founder anecdotes. It is what both of them point at when you read them carefully.
Sorkin’s combination is worth stealing for the generative half specifically. He does not only walk, he walks and speaks the ideas aloud, and per Lound the pairing compounds because hearing your own idea creates a small distance that pure silent thinking does not. For a founder, that translates to a walking meeting with one other person, talking the problem out, rather than a solitary walk where the idea never leaves your head. The output of a Sorkin walk is dialogue. The output of a founder’s walk should be a short list to take back to the desk.
The thing to notice is how rarely the famous accounts contradict the data. Jobs walked for serious conversations and brainstorming with Ive, not to close the books. Sorkin walks to write dialogue, the most divergent task there is. Merchant walks to reframe problems, then sits to act on the reframing. The study found walking helped 81 percent generate ideas and 23 percent solve closed problems, and the people who built the habit without ever reading it ended up using it for the first thing and not the second.