Decide when and how you will do it
Choosing a promising activity still leaves a practical question: when and how will it happen? “We should walk together sometime” leaves both friends to make several decisions later. “Meet outside the office on Tuesday at 12:30” settles the first few decisions, provided that time and place actually work for both of them.
Think through what would make the walk possible. The friends need overlapping breaks, a route they can both use, and enough time to eat. If one person is expected to answer calls throughout lunch, putting a walk in the calendar will not remove that obligation. They need to find another time or choose a different activity.
Changing the conditions can make an activity easier to do. You might choose a closer meeting point, get the equipment ready, set aside enough time, or move the activity to another day. Plan around the responsibilities you already have, so that the activity has a place in an ordinary week.
What a specific plan can add
A field experiment by Katherine Milkman and colleagues gives a useful example. More than 3,000 eligible employees at a US utility received information about free workplace flu-vaccination clinics. Some mailings also asked employees to write down the date and time they planned to attend. The basic opportunity to get vaccinated was already available to all groups.
The group asked to plan a date and time had a vaccination rate of 37.1%, compared with 33.1% in the control group. That is an observed difference of four percentage points. The result shows what this prompt achieved at these clinics. It does not tell us that making a detailed plan will help with every behavior-change problem.
Notice what the prompt did not provide: the clinic, the vaccine, or an accessible location. Those conditions were already in place. For our friends, choosing a meeting time may be useful once they know what they both want to do and whether their lunch breaks overlap. It cannot create a lunch break neither person has.
Milkman and colleagues (2011): Planning prompts and workplace vaccination
Record whether it happened and whether it helped
Keep track of two things: whether people did the activity and whether it helped them get what they wanted. Otherwise, you could spend time persuading people to follow a plan without noticing that the plan itself is not helping.
Suppose the friends complete a walk but spend most of it answering work messages. “Walk completed” is an accurate record of the activity. It is an incomplete record of the result, because the conversation they wanted barely happened. They might keep the walk but change the time, or decide that another setting gives them a better chance to talk.
A useful note could be as simple as: “We met, walked for twenty minutes, and caught up until a work call interrupted us. We both liked it, but Tuesday lunch may be unreliable.” The note says what they did, how it went, and what got in the way. That is enough to help them decide what to try next.
In an organization, you may need a more consistent record. A team testing a new handoff procedure could track whether requests received an owner and whether fewer requests were left unanswered. Recording only completion of the procedure would miss the reason for introducing it.
Make room for effortful practices
As an activity becomes familiar, some parts may require less conscious deliberation. Habit research describes how repeating a response in a recurring context can support automaticity: the context increasingly prompts a learned response. That helps explain why an ordinary cue can set a familiar action in motion.
But starting an activity is different from doing the work itself. Opening a document at a familiar time can become routine; deciding how to explain an idea, organize an argument, or repair a confusing paragraph still requires judgment. Getting used to opening the document does not make those decisions for you.
Exercise research has examined this distinction directly. In an observational study of 123 adults, Phillips and Gardner separately measured the habit of initiating exercise and the habit of performing an exercise routine. When the researchers considered both together, people who got started more habitually tended to exercise more often. Once that was taken into account, knowing how habitual their exercise routine was did not help predict how often they exercised. Because participants were not randomly assigned to develop these habits, the study does not prove that helping someone make the start habitual would cause them to exercise more. It does show why treating the entire activity as one habit can obscure a useful distinction.
This distinction is central to what Jason means by habits and practices. The practical mistake is to expect complex work to become effortless simply because you repeat it. If the activity still requires attention, you may need to protect time, develop skills, or choose work that holds your interest. A better cue cannot do all of those jobs.
For the friends, remembering to leave the office is only the beginning. Listening, asking questions, and responding to what the other person says remain part of the conversation. They need time to pay attention to each other, not just a reminder to turn up.
Wood and Rünger (2016): Psychology of Habit · Jason’s Habits vs. Practices · Phillips and Gardner (2016): Exercise initiation and execution habits, original manuscript
Use a personal trial for the questions it can answer
After a few attempts, the friends can learn whether their breaks overlap in practice, whether they enjoy the route, and whether they have enough time to talk. Those observations can help them choose what to do next. They do not establish that walking causes better friendships in general.
Even a before-and-after improvement leaves other explanations open. Perhaps a deadline ended during the same week, so both friends became less distracted. Perhaps they had more to talk about because one had just returned from a trip. Without a suitable comparison, they cannot separate the contribution of the walk from everything else that changed.
That limitation does not make the experience useless. If both enjoy the walks and have time to talk, they may have enough reason to continue. The evidence needed for that modest decision is different from the evidence needed to sell a walking program as a proven way to improve relationships.
Match the comparison to the size of the decision
An organization deciding whether to introduce a costly program has more at stake. Imagine customer response times improve after a new handoff procedure begins. If the team also hired two people and received fewer requests that month, the improvement cannot confidently be attributed to the procedure alone.
A randomized comparison can help by assigning eligible people or units to different conditions rather than letting the groups form through preference or convenience. The outcome then needs to be measured in the same way over the same period. The appropriate unit might be a customer, a team, or a site, depending on how the intervention works and whether people in one group affect the other.
A small pilot can reveal confusing instructions or missing resources without being large enough to estimate an effect reliably. Before scaling a program, decide what size of improvement would justify it and what study would distinguish that improvement from ordinary variation. More participants, a meaningful comparison, or longer observation may be needed; the right choice depends on the claim you want to make.
Decide what to keep, change, or stop
Choose a point at which you will review the trial. The timing should fit the activity and the outcome: you may discover on the first walk that lunch is too rushed, while other benefits may take longer to notice. Choose a time to review that gives you a fair chance to learn what you need to know.
If the activity helps and remains workable, continue. If you can remove a specific obstacle, make that change and see whether it helps. If the activity keeps falling short of what you wanted, return to the menu and try an alternative. You can also stop when the cost is no longer justified by the benefit.
For the friends, repeated work interruptions might be a reason to move the walk to a different time. If there is no shared lunch break, the call option may be more sensible. The decision becomes clearer when they ask what gave them time together rather than how faithfully they followed the first plan.
A trial plan you could actually use
The friends agree to try a twenty-minute walk during a shared Tuesday lunch break. They choose a meeting point and check that the route suits both of them. After each attempt, they note whether they met, whether they had time for a conversation, and any obstacle that made the arrangement difficult.
They decide to review after three attempts, when they expect to know more about whether Tuesday lunch works for them. If they enjoy the walks and can make time for them, they will continue. If lunch repeatedly falls through, they will try a call at another time. They can use their notes to choose a time or activity that works better, rather than blame themselves for missing a walk.
Check your understanding
A team answers customer requests faster after introducing a new procedure. It also hired two people that week. What can it conclude about the procedure?
Response times improved while the procedure was in use. The team cannot yet tell how much the procedure contributed, because additional staff may have helped too. A suitable comparison would help separate those explanations; observations alone may still reveal whether the procedure is clear and practical.
Try it yourself
Use the behavior-menu exercise to record your chosen activity, when and where it could happen, and what needs to be ready. Decide how you will tell whether you did the activity and whether it helped. Then choose when to review what happened and describe what would lead you to continue, change the conditions, switch activities, or stop. Keep the plan somewhere you will see it when you need to use it.
Read further
The five lessons bring you back to a practical question: which behavior serves this goal for these people, under these conditions? Your first trial will give you more information for answering it. Keep the goal clear and use that information to improve the next choice.