If a behaviour requires you to be at your best to happen at all, it will happen only on your best days — which, over a year, is not many days. This is the central and slightly deflating fact about habits. The reliable move is almost never to become a more disciplined person. It is to change the conditions so that less discipline is required.
That reframing is worth taking literally. Treat the behaviour you want as a system with inputs and obstacles, and go looking for the obstacles the way you would in anything else you were trying to make work.
Count the steps
Write down every physical action between where you are now and the behaviour starting. Not the intention — the actions. Find the running shoes, find socks, change, find the key, leave. Open laptop, find the file, close the seventeen tabs, start.
Every step is a place the sequence can stop, and the stopping is not conscious. It is not that you decide against running at step three; it is that step three is where the sequence loses momentum and your attention gets picked up by something else. Removing two steps does more than a week of resolutions.
Defaults beat decisions
Anything you have to decide fresh each time will eventually be decided against, because deciding is expensive and you are frequently tired. Anything that happens unless you intervene will mostly happen.
So the useful question is not "how do I remember to do this" but "what would have to be true for this to be what happens by default". Clothes already out. The document already open on the screen you will see first. The standing block already in the calendar so that declining it requires an act. The walk being the route home rather than an addition to it.
Ask what happens if you do nothing at all. If the answer is the behaviour you wanted, you have designed it correctly.
Make the unwanted path expensive
The mirror of the same idea. You are not going to out-argue a habit at the moment of temptation, because at that moment the part of you doing the arguing is the part that is losing. Raise the cost in advance, while you are calm and it is cheap to do.
The specifics are unglamorous and effective: the phone charging in another room, the app removed rather than resisted, the snack not in the house rather than in the house being ignored, the browser signed out. None of this is a statement about your character. It is an acknowledgement that in-the-moment willpower is a poor and inconsistent mechanism, and that you have a much better one available at a quiet moment on a Sunday.
One object, one meaning
Physical spaces carry instructions. A desk used for work, bills, meals and television gives an ambiguous signal, and ambiguity costs you the first ten minutes every time. A single object or place reserved for one behaviour becomes a cue that requires no self-talk: this chair is for reading, this notebook is only for the morning pages, this table is not where the laptop lives.
The reservation only works if it is respected, which is why one reserved thing is better than a whole redesigned room. Small and honoured beats large and eroded.
Attach to what already exists
New behaviours have no place in the day, so they have to find one, and there is no free space. Chain the new thing to an existing reliable event rather than to a time: after the kettle boils, after the last meeting ends, after the school drop-off, immediately on sitting down. The existing event does the work of remembering, and you get to skip the whole category of failure that consists of simply not thinking of it.
Expect the design to decay
Arrangements erode. The shoes migrate, the calendar block gets colonised, the phone comes back to the bedside. When a habit lapses, the instinct is to interrogate your motivation; the better first move is to go and look at the physical set-up and find out what changed. Nine times out of ten something did, and putting it back is a five-minute job that no amount of self-examination would have produced.
This is also why habit work is never finished in the way people hope. It is maintenance of an arrangement rather than the acquisition of a virtue — which is worse news for the story and much better news for the odds.