Label the cue and routine; keep the reward as a hypothesis

The cue is the event or condition that precedes the behavior. The routine is what you do. The reward is what the behavior provides or is expected to provide. Duhigg's practical method involves observing the routine, experimenting with rewards and examining the cue before making a plan.

A neat table can make the explanation look more certain than it is. If you check your phone after a difficult paragraph, difficulty may be a cue, but the room, time or a notification could also matter. If checking feels relieving, that observation still does not prove the same reward drives every episode. Write down what happened before deciding why.

Three original examples to practise with

1. Chat checking. While drafting a document, a preview appears on your phone. You unlock it, read the chat and keep browsing. You observe the preview and the sequence of actions. A possible reward is finding out what someone said; another is escaping a difficult sentence. On a later occasion, you might check only the relevant message or take a short planned break without opening other apps, then note what remains unsatisfied.

2. Walking to a colleague's desk. Around mid-afternoon, you leave your screen and start a brief conversation. You might be seeking company, movement, help with a task or a pause. Compare a short walk alone with a conversation at another convenient time, while respecting the colleague's availability. Record whether the urge to leave the task changes after each.

3. Post-dinner video browsing. After clearing the table, you sit in the same chair and open a video feed. Perhaps the reward is entertainment, quiet time or a transition away from chores. Try one chosen short video with a clear stopping point and observe whether it meets the need for a break. A reading session may serve a different purpose, so compare alternatives you enjoy.

ExampleObserved cue candidateObserved routinePossible reward to test
Chat checkingA message preview during draftingUnlock, read and continue browsingInformation, contact or relief from difficulty
Desk conversationA mid-afternoon pause in workWalk over and start talkingCompany, movement, help or a break
Evening browsingSitting down after clearing dinnerOpen the feed and keep watchingEntertainment, quiet or a transition

Record a few episodes before choosing a change

Keep a short note of the time, place, preceding event, action and what you noticed afterward. You can add your mood and who was present if those details seem relevant. Use ordinary descriptions: 'I opened chat after failing to finish a sentence' is clearer than 'My reward system malfunctioned.'

Compare several episodes. Did the same preceding event recur? Did you feel differently after a conversation than after solitary browsing? Your notes may reveal more than one pattern. That is useful information, even if it prevents a simple one-line loop. Do not change the explanation merely to make every episode fit the first guess.

Test one possible reward with a small alternative

Suppose you think afternoon checking provides a break. On one suitable occasion, step away from the screen briefly, then return without opening the feed. Record whether the urge remains and whether you can resume the task. On another occasion, if you need information, check the specific message. Different results may help distinguish what you were seeking.

Change one element at a time where practical. A single episode does not prove the cause, and a busy day may differ from a calm one. If the alternative leaves the urge unchanged, record that result and test whether you were seeking something else.

Make a plan that includes the actual situation

Write a specific plan after observing: 'When the afternoon urge appears during drafting, I will save the sentence, take a short break and return to the next written step.' Decide how an urgent message will be handled. A plan that ignores genuine communication needs may fail for reasons unrelated to willpower.

After a few attempts, review the cue, the alternative routine and its result. Keep the parts that serve your goal and adjust the rest. This lightweight worksheet is for ordinary behavior experiments. Complex or harmful patterns may require support beyond a self-directed habit map; the three-part model alone cannot explain every difficulty.

Check your understanding

1. A message preview appears, and you unlock your phone to browse. Which part is the routine?

Answers and explanations
  1. The message preview appearing

    This is an observed cue candidate: it precedes the behavior.

  2. Unlocking the phone, reading and continuing to browse (Correct)

    This is the sequence of actions you perform after the cue.

  3. The assumption that you wanted a break

    This is a possible reward hypothesis, not the observed action.

2. You think an afternoon desk visit provides company. What would help investigate that hypothesis?

Answers and explanations
  1. Observe several episodes and compare a suitable conversation with a short solitary break (Correct)

    Comparing what follows different alternatives can help you examine whether contact, movement or another need matters.

  2. Declare the reward certain after writing it in a table

    A table organizes a hypothesis but does not establish its cause.

  3. Assume every colleague wants a conversation whenever you do

    Your plan needs to respect other people's availability and cannot infer their preferences.

3. An alternative routine did not reduce the urge on one occasion. What is the most reasonable next step?

Answers and explanations
  1. Conclude that no habit can ever change

    One result cannot support a universal conclusion about all behavior.

  2. Hide the result so the original hypothesis stays tidy

    The unexpected result is information you need to refine the explanation and plan.

  3. Record the conditions and consider whether a different need or cue was involved (Correct)

    You use the observation to update the hypothesis rather than treat the initial map as proven.

Map and test one everyday pattern

Your notes stay on this page and disappear when you reload it.

Sources and further reading

An original KeenRead guide with new examples and questions. Book names identify the source works; KeenRead is independent of their authors and publishers.

Try a narrated idea

Read an idea, answer a question, and see the explanation in KeenRead’s free web sample.

Try a narrated idea