What does it mean that craving engineering?

You can create cravings for positive behaviors by consistently pairing them with rewards.

Example: In the early 1900s, almost no Americans brushed their teeth. Claude Hopkins, the advertising pioneer, was hired to promote Pepsodent toothpaste. He did not succeed by explaining dental hygiene. He succeeded by engineering a craving. Hopkins told people to run their tongue across their teeth and feel the film — a natural coating that had always been there but that no one had noticed as a problem. That was the cue. Then Pepsodent added citric acid and mint oil to the formula — ingredients that served no cleaning function but created a cool, tingling sensation in the mouth. That was the reward. Within a decade, toothbrushing rates in America went from roughly 7 percent to 65 percent. People did not start brushing because they understood plaque. They started brushing because they craved the tingle. The craving did not exist before Hopkins and the Pepsodent chemists manufactured it.

Try this: Choose one behavior you want to perform daily but currently have no craving for — a behavior that you know is beneficial but that generates no anticipatory pull. Design a craving engineering protocol for it using all five steps from this lesson. Step 1: Define the cue — a specific time, location, or preceding action that will trigger the behavior. Step 2: Design the reward — choose something immediate, sensory-rich, and genuinely pleasurable that you will pair with the behavior every single time. Step 3: Write out the full pairing sequence — what happens at the cue, what the behavior looks like, and how the reward is delivered within seconds of completion. Step 4: Commit to running this pairing for thirty consecutive days, logging each instance and rating (on a 1-5 scale) how much anticipatory pull you feel at the cue before performing the behavior. Step 5: After two weeks, review your ratings. If anticipatory pull is increasing, the engineering is working. If it is flat, redesign the reward — it is not salient enough to generate a dopamine prediction.