You may be solving the wrong problem
Suppose you want to see your friends more often. You decide to organize a dinner every Friday, create a group chat, and send out a scheduling poll. One friend cannot do Fridays. Another is traveling. A third likes the idea but never replies. Three weeks later, you are looking for better reminders and wondering why getting together takes so much effort.
At this point, it is easy to assume that the problem is getting people to commit to dinner. But dinner was only your first idea. What you wanted was time with your friends, and there may be ways to have that time without coordinating an entire group on a busy evening.
A walk with one friend, coffee near work, or a phone call during a commute might serve the same purpose. These options are not interchangeable: a call will not give you everything a dinner does. The important question is which parts of the experience you actually want, and which arrangements make those parts possible.
We will return to this imagined friendship example throughout the five lessons. Alongside it, we will look at other situations so you can see how the same reasoning applies to learning, work, and the design of products or services.
Separate the result from the route
An outcome is the result you want to bring about. A behavior is something a person does that could help bring it about. In the friendship example, the desired result might be feeling more connected because you have time to talk properly. Hosting dinner is one possible behavior; arranging a walk is another.
To see why the distinction matters, imagine that the Friday dinners become a success on the calendar. Everyone attends, but the restaurant is so loud that you can barely hear each other. If you measure only attendance, the plan looks excellent. If you return to the reason for meeting, you can see that something important is missing.
Now imagine the opposite: you stop organizing dinners but regularly have good conversations with friends in other settings. You have abandoned the original activity without abandoning the result. Keeping those two things separate lets you change your approach without treating every change of plan as a failure.
Try stating your outcome without naming the app, product, or routine you first had in mind. Instead of “use my language-learning app every day,” you might write “understand enough spoken Spanish to follow a conversation with my relatives.” That wording gives you a way to judge the app, as well as a reason to consider listening practice, conversation, or a class.
Make sure your measure reflects the benefit
Organizations face the same problem, often with better dashboards. Imagine a support team that wants customers to resolve problems quickly. The team launches a help app and makes daily app use its main measure of success. More visits look encouraging, but they could mean that customers find the app useful—or that customers have to keep returning because their problems remain unresolved.
App use is a proxy: an observable measure being used to stand in for something else. A proxy can be useful, but it needs a defensible connection to the result. Here, the team would also want to know whether customers solved their problems, how long that took, and whether they had to contact support again.
The lesson is not to stop counting activities. Count them when they tell you something useful, then check the benefit they are supposed to produce. If a clearer instruction page lets customers solve a problem in one visit rather than four, fewer visits could be good news.
Personal goals rarely need a dashboard, but they need the same care. If your goal is a closer friendship, asking whether you had an unhurried conversation may be more informative than counting invitations. Decide what you would notice if things improved, even when the answer is partly a judgment rather than a number.
A research example: studying versus remembering
Suppose a student wants to remember a passage a week after reading it. Rereading is one possible activity. Closing the passage and writing down what they can recall is another. The second activity is called retrieval practice because it requires bringing information back from memory.
Roediger and Karpicke compared these activities in a 2006 experiment with 120 university students. When tested after five minutes, students recalled more after rereading. When tested after a week, they recalled more after practicing recall: an average of 56% of the main points in the passage, compared with 42% after rereading.
The important point is that the better activity depended on when learning was measured. A method that produced better immediate recall did not produce better delayed recall. The study concerned memory for short passages, not every kind of learning. But it makes our distinction concrete: choose the outcome—including when you need it—before judging the activity.
Roediger and Karpicke (2006): Test-enhanced learning, original paper
Check whose problem you are solving
Before choosing an activity, find out whether the people involved want the result you are asking them to help produce. A manager may want daily status updates because missed handoffs are delaying work. Employees may want fewer interruptions and clearer responsibilities. There may be a shared problem here, but a daily update is not automatically the best solution for either group.
A useful conversation would ask where handoffs actually fail. Perhaps no one knows who owns the next step. If so, assigning an owner when work changes hands might help more than asking everyone to write another report. The point of asking employees is to learn what the proposed behavior would require of them and whether it addresses the difficulty they encounter.
This also applies to your own goals. You can want to have written a novel while disliking the process of writing one. That does not settle whether you should write, but it is a real demand to account for. A plan based on becoming a different person who suddenly loves every part of the activity leaves an important question unanswered: what will make the work worth returning to for you?
Give yourself more than one way forward
Once you have described the result, list several actions that could help produce it. Look for genuinely different routes before making small improvements to the first one. Three reminder schedules still depend on the same dinner; a walk, a call, and a dinner place different demands on the people involved.
Jason calls a list of possible activities for a goal a Goal Menu. Its value at this stage is practical: it keeps alternatives visible so that choosing a behavior becomes part of your work. Later lessons will show how to compare the options by their likely usefulness, their demands, and the person who will do them.
You do not need to delay every ordinary decision while you make a comprehensive list. But when you have repeatedly struggled to make one behavior happen, reconsidering the behavior itself is a useful place to start.
From a dinner plan to a friendship goal
The friends replace “organize dinner every Friday” with “have regular chances to catch up without rushing.” They keep dinner as an option and add a lunchtime walk and a phone call. Each option now has to answer the same question: will it give them the kind of conversation they are missing?
A brief call might suit two friends who mainly want to hear how the other is doing. It might be a poor substitute for a friend who wants a long evening together. Clarifying the outcome does not tell everyone to choose the easiest activity; it tells them what the activity needs to accomplish.
Check your understanding
A support team wants customers to open its help app every day. Has it named the benefit it wants to create?
It has named an activity. To judge whether that activity helps, the team still needs to specify the benefit—for example, customers resolving their problems quickly. More app visits could mean useful help, but they could also mean repeated attempts to solve an unresolved problem.
Try it yourself
Choose a goal you care about and write what you want to be different. Then name the activity you have been treating as the solution. Describe how you would recognize progress toward the goal, and list at least two other activities that might produce it. For each alternative, explain the connection to the result rather than simply calling it a good idea.
Read further
You now have a result to aim for and more than one possible route. The next lesson asks what evidence would give you a reason to expect an option to help.