Friction as the Master Lever

Most behavior change fails because it asks more of a person, and the person is tired. Choice architecture asks less of the room instead. This page makes the case that the number of steps between you and an action — friction — is the most reliable dial you own, then turns it into a checklist you can apply tonight.

The Step Count That Decides

Friction is simple to define and easy to forget: it is the number of steps between you and an action, counted in time, distance, decisions, and fuss. Open a drawer, find the lid, dig for the pan, wash the pan — a cooking habit just failed before it started. Pick up the phone, unlock it, tap the icon — a scrolling session just started without a decision at all. The two columns of that arithmetic are what this page is about: the good path should have fewer steps than the bad one, and most of the time in real households it is the opposite.

The reason friction outperforms motivation is structural: motivation is a wave — highest at eight in the morning, lowest at ten at night — and exactly when the wave is lowest, your environment makes its decisions. Friction never gets tired. A lid on the cookie jar does the same job at 8 a.m. and 10 p.m., on a good or bad day. The lever multiplies whatever force you have, including the near-zero force of a Tuesday night.

Where the Evidence Lives

The cleanest demonstrations come from food, because everyone repeatedly chooses it in full view. In one classic office study, office workers ate several more candies a day when the same dish sat clear and open on their desks than when it sat two meters away in a drawer — same candy, same people, only the arrangement changed (Wansink, Painter & Lee, 2006). An earlier study found the same pattern with a transparent bowl versus an opaque one — seeing the food mattered as much as tasting it (Painter et al., 2002). The honest read: the environment supplies a running vote on every choice, tallied in steps.

The principle extends past snacks. Wendy Wood's work on habit shows that behavior becomes bound to context — the cue is the situation, not the goal — and that changing the situation changes what the cue triggers (Wood & Rünger, 2016). The famous demonstration is the stale popcorn: moviegoers with strong theater-going habits ate stale popcorn at almost the same rate as fresh, because the context alone pulled the behavior; people without the habit noticed the difference immediately. Context, not taste, was running the show. The parent protocol page documents this line of evidence in full; this page applies it as engineering.

Subtract First: The Good Path

Start by removing steps from the behavior you want, while you are motivated — the work happens once, at high energy, so the low-energy moment inherits a ready-made path.

• 👟 Lay the kit out the night before. Running clothes on the chair by the bed deletes the 6 a.m. negotiation over where everything is. The decision was made at 9 p.m. by the person who was not tired yet.
• 💧 Put the water bottle on the desk, filled. Hydration that requires standing up, walking somewhere, and remembering loses to hydration that is already there. One step removed is often the difference between a habit that runs and one that does not.
• 🎒 Park the packed bag by the door. Gym bag, work bag, hiking bag — packed the evening before and left at the exit. The door has a habit of becoming the last excuse; remove the excuse, keep the door.
• 📚 Put the book on the pillow. The evening default is whatever is closest to the bed. Make it the book, and move the phone to the charger elsewhere in the house.
• 🥗 Put the vegetables at eye level. The first thing seen in the fridge tends to be the first thing eaten. The crisper drawer is a place food goes to be forgotten; the eye-level shelf is where food goes to be eaten.

Add Second: The Bad Path

Then add steps to the behavior you want less of. The goal is not prohibition — it is making the unwanted behavior cost more than the wanted one, so the default choice shifts without a daily decision.

• 🔒 Put the snacks in opaque bins on the top shelf. Out of sight and two steps up. The studies say the effect is real: invisibility and inconvenience together cut intake far more than either does alone.
• 🗂️ Move the feed apps behind a search. Delete them from the home screen and force a deliberate trip to find them. The default scroll now requires a decision, and decisions at 10 p.m. are expensive.
• 🔕 Turn off autoplay everywhere. The next episode should never load itself. Every auto-loaded video is a step you were not asked to take; reclaim it.
• 📱 Charge the phone outside the bedroom. One move that adds a step to the midnight scroll and subtracts a step from falling asleep — the Bedroom Engineering protocol page documents the sleep side of this swap.
• 🚪 Keep the big-ticket temptations out of the house. The grocery store is a harder place to resist a bottle of wine than the kitchen counter. Buying it once a week and keeping it out of reach beats refusing it every night.

The do / avoid table in four rows — restraint not required, arrangement required.

The Exchange Rate

The arithmetic: one step off the good path and one onto the bad path beats a doubled dose of motivation. Willpower depletes across the day; friction does not. The chart plots the practical version — how likely a behavior actually is as its step count climbs. It is an illustrative curve, not a measured one, but every point on it is consistent with the candy-dish findings and the context-dependence work cited above.

Practical Rules

Four rules keep the lever aimed correctly over months, not just the first enthusiastic week.

• 1️⃣ Change one path at a time. Pick one good habit and one bad habit; adjust their steps and nothing else for a week. Ten half-installed systems fail; one complete swap tells you whether the lever is working for you at all.
• 2️⃣ Make the change stupidly big. Half-hidden candy still gets eaten; apps that take two taps still get tapped. If the step costs less than the impulse, the step loses. Overshoot the friction on the bad path and undershoot it on the good one.
• 3️⃣ Reversible is fine — reversible is good. You are not signing a lease. A rearrangement that does not work costs ninety seconds to undo. Treat every change as an experiment with a clear yes/no question.
• 4️⃣ Pair the environment with a cue. Friction lowers the cost of a habit; it does not trigger it. The trigger comes from an if-then plan — the Implementation Intentions protocol page builds that half of the machine, and the default effect page in this series builds the do-nothing option.

Related Topics