Why does friction engineering behavior change fail?

The most common reason friction engineering behavior change fails: Treating friction engineering as a one-time fix rather than an ongoing calibration. You rearrange your environment once, the behavior improves for two weeks, and then you slowly undo the friction: you bring the phone back to the nightstand because you need the alarm, you reinstall the app because you need it for just one thing, you move the junk food back to eye level because you had guests. Every friction reduction you introduced gets eroded by convenience drift. The fix is to schedule a monthly friction audit — walk through your environment and check whether the friction gradients still point in the right direction. Friction engineering is not an event. It is maintenance.

The fix: Run a friction audit on one behavior you want to do more of and one behavior you want to do less of. For each, list every micro-step between the impulse and the action. Count them. Then redesign both: remove at least two steps from the desired behavior and add at least two steps to the undesired behavior. Implement the changes physically — move objects, change app layouts, rearrange your space. Track the behavior for seven days. Do not rely on willpower at any point. If you find yourself needing willpower, your friction engineering is not aggressive enough. Redesign and try again.

The underlying principle is straightforward: Adding friction to bad choices and removing friction from good choices changes behavior.