You have run the same loop for three years. Same streets, same playlist, same pace within a minute or two. You are proud of the streak, and you should be. But last spring a neighbor started running, and last week they passed you on the hill and did not look back. Nothing about your effort changed. Something about its effect did. There is a particular quiet shame in doing something faithfully and watching it stop giving anything back, and almost nobody talks about it, because the advice we are given is all about starting and sticking, never about what happens after you do.

The plateau that arrives right after success

Psychologists Paul Fitts and Michael Posner described skill learning as moving through three stages. In the cognitive stage, everything is deliberate and clumsy. You think about each footfall, each chord, each sentence. In the associative stage, the pieces start linking up and errors drop. Then comes the autonomous stage: the skill runs without attention. You can do it while thinking about dinner.

Every habit guide, including this blog, treats that third stage as the finish line. For habits, it is. The whole point of a habit is that it no longer costs you a decision. But the autonomous stage has a property that goes unmentioned. Once a behavior runs on its own, your brain stops paying attention to it, and a brain that is not paying attention is not learning. Repetition keeps the skill where it is. It does not push it anywhere.

The writer Joshua Foer, drawing on the work of psychologist K. Anders Ericsson, called this the "OK plateau": the level at which you are good enough that performance becomes automatic, and improvement silently stops. Most of us live on dozens of these plateaus. We type at the speed we typed at in our twenties. We cook the twelve meals we cooked a decade ago. We journal the same three paragraphs about the same three worries.

Why doing it more doesn't make you better

Ericsson spent his career studying expert performers, and one of his more uncomfortable findings was that experience alone is a poor predictor of expertise. Beyond a certain point, people with decades of practice are frequently no better than people with far less. A widely cited systematic review of physicians, published in the Annals of Internal Medicine, found that more years in practice were often associated with worse performance on measures of medical knowledge and quality of care, not better. These were people doing the thing every single day.

The mechanism is not laziness or decline. It is automaticity working exactly as designed. When you first learned the skill, your attention was fully engaged, and attention is what drives the adjustments that produce improvement. You noticed errors and corrected them. Once the skill ran on autopilot, that feedback loop closed. You stopped noticing the errors because you stopped noticing anything. The behavior got efficient at the cost of getting any better.

Ericsson's alternative, which he called deliberate practice, has a few defining features. It targets a specific aspect of performance that is just beyond your current ability. It involves full concentration rather than autopilot. It gives you immediate feedback, so you know whether the attempt worked. And it is, by his own repeated description, not enjoyable in the moment. You are deliberately walking yourself back out of the autonomous stage into the clumsy cognitive one. That is why it works, and why almost nobody does it voluntarily.

Foer's own example was typing. People plateau at a typing speed and stay there for decades, because typing at that speed is comfortable. The way to improve, he found, was to force yourself to type faster than you comfortably could, accept the errors that caused, and pay attention to exactly where they happened. The discomfort was the signal that learning had been switched back on.

Habits and skills obey different rules

Here is the one idea worth taking from all this. A habit and a skill are not the same thing, and they want opposite things from you.

A habit's goal is automaticity. You want brushing your teeth, locking the door, and showing up to the gym to require no thought at all. For these, the plateau is the point. Trying to "improve" at flossing would be absurd.

A skill's goal is growth, and growth requires the opposite of automaticity. It requires attention, discomfort, and feedback. If you treat a skill like a habit, you will get exactly what habits deliver: reliable, effortless, unchanging performance. Which is wonderful, right up until it is not what you wanted.

Most people never decide which one they are building. They start running to "get fit," and the habit of running forms beautifully, and then they are confused and a little ashamed when their fitness stops changing. The habit succeeded. The skill was never asked for. The same thing happens with the piano, the second language, the writing, the cooking. The consistency is real. The expectation that consistency alone would keep producing change was the error.

The reframe that helps is to split the thing in two. Keep the habit exactly as it is: the showing-up, the cue, the time, the automatic part that gets you to the door. Then, inside the habit, carve out a small, deliberate slice where you are not on autopilot. The habit carries you to the practice. The practice is where the growth happens.

The small uncomfortable step

Cognitive psychologist Robert Bjork has spent decades showing that the conditions which make learning feel hard, which he calls desirable difficulties, are often the conditions that make it stick. Spacing practice out, mixing problem types, testing yourself instead of rereading. Each feels worse in the moment and produces more durable learning. The feeling of ease that comes with the autonomous stage is pleasant precisely because nothing new is being encoded.

This matters for anyone on a plateau because it reverses the usual instinct. When a habit starts to feel easy, we read that as success and keep doing more of the same. Bjork's work suggests reading it as a cue. Ease means the learning has stopped. If growth is what you want, the next move is to introduce a small, specific difficulty that makes the skill feel slightly clumsy again.

Small is the operative word. The stretch that works is just past your current edge, not a leap. Ericsson's expert performers did not practice at the limit of exhaustion. They practiced at the limit of their attention, on one narrow thing, for bounded periods, and then they stopped. You do not have to make the whole habit harder. You have to make a few minutes of it conscious.

Your next moves

  • Sort your habits into "keep automatic" and "want to grow." Write down everything you do regularly and mark each one. Teeth, dishes, the commute: automatic, leave them alone. Running, writing, the instrument, the language: these are skills, and they need a different plan.
  • For each "grow" item, name one specific thing you are currently bad at. Not "get better at running" but "hold pace on the hill." Not "improve my writing" but "stop opening every paragraph the same way." You cannot practice deliberately at something you have not named.
  • Carve a ten-minute deliberate slice inside the existing habit. Keep the cue and the routine untouched. Inside it, spend ten minutes with full attention on the named weakness, at a difficulty where you fail some of the time. Then return to autopilot for the rest.
  • Build in feedback you cannot avoid. Record yourself playing. Time the hill. Reread yesterday's paragraph aloud. Automaticity survives on not noticing, so arrange to notice.
  • Expect it to feel worse before it feels better. The clumsiness is the cognitive stage returning. If the slice feels smooth and pleasant within a week, it is probably no longer a stretch, and it is time to nudge it again.

Where Cadence fits

Cadence is built around one belief: small steps, big change. Most of the time that means keeping a step small enough that you will actually take it. But the other half of "small" is this: once a step has become automatic, the next small step is the one that makes it slightly uncomfortable again. Cadence lets you hold the habit steady while you adjust the step inside it, so that showing up stays effortless and the few minutes that matter stay deliberate. If you have been faithful to a habit that has stopped giving anything back, that nudge may be all that was missing. Start with one step at https://cadence.lumenlabs.works.