The usual explanation for a failed habit is that you did not want it enough. That is almost never the real story. Far more often the habit was doomed at the moment you chose it — wrong behaviour, wrong hour, wrong week, decided in about four seconds while feeling optimistic.
A pact fixes the follow-through problem. It does not fix a bad pick. So before you agree to eight weeks of anything with another person, spend one week finding out what your days are actually made of.
You are not designing on a blank page
Your week is already close to full, and most of what fills it is automatic. Researchers who had people keep hourly diaries found that a large share of everyday behaviour — somewhere around two-fifths — is performed almost daily, in the same setting, while the person is thinking about something unrelated.
That is the part worth sitting with. Automatic behaviour does not announce itself. It does not feel like a choice you are making; it feels like the shape of the day. Which means that when you slot a new habit into 7am, you are not filling empty space. You are competing with something already there that you have never consciously looked at.
Log one week, four columns
The audit is small on purpose. An elaborate tracking system abandoned on Wednesday teaches you nothing.
For seven days, write down anything that took more than a few minutes. Four columns: what it was, when, how long, and what set it off. That last column is the whole exercise. Did you decide to do it that morning? Did it happen because you were already standing there? Did someone else's schedule pull you into it?
Do not tidy the log. The days that went badly are the informative ones, because those are the days your pact will have to survive.
Three things the log almost always shows
- Your best intentions have no trigger. The habits you keep failing at usually live in an unclaimed slot — "in the evening", "at some point". Nothing fires them. The habits that run without effort are pinned to something concrete: leaving the office, the kettle boiling, the kids going to bed.
- One hour is already carrying too much. People pile new habits onto the hour they wish were free rather than an hour that is. If 6:30am already holds the commute prep, it will not also hold thirty minutes of exercise, no matter who you promised.
- You already have a habit nobody credits you for. Almost everyone finds at least one genuinely consistent behaviour in the log — a walk, a stretch, a weekly call. It is worth noticing, because it is the anchor for everything below.
Attach, do not invent
Here is the payoff. A new behaviour needs a cue, and manufacturing a cue from nothing is the hardest available route. Your log is a list of cues that already fire reliably, at known times, without you doing anything to maintain them.
So pin the new thing to one of them. Reading goes after the kettle, not "before bed". The walk goes immediately after the last meeting ends, not "at lunchtime". Half the automation you need already exists — you are borrowing it rather than building it.
This also tells you which habit to pick first. Given two candidates you care about equally, take the one that can hang off a cue you already have. It will still be running in week six.
Then set the bar at your worst day
The log gives you something else useful: an honest picture of a bad week. Look at the two or three worst days in it and define the habit at a size those days could have absorbed. Ten minutes. One set. One page.
People consistently write pacts for the version of themselves who had a good Tuesday, then break them on a Thursday that went sideways. A commitment your worst day can clear is not a lowered standard — it is the only kind that produces a long run of them.
Turn the audit into the pact
Bring three things from the log to whoever you are pairing up with. The cue you are attaching to, stated as an actual moment in the day. The floor version of the behaviour, defined precisely enough that neither of you could argue about whether today counted. And the days the log says are genuinely at risk, so the skip allowance is set with those in mind instead of being negotiated in the moment.
Then re-run the audit whenever your context changes. A move, a new job, a different shift pattern — habits are bound to their setting, so a change wipes them and silently installs replacements. That week is the best window you will get to choose what goes in the slot, and it is the right week to renew the pact rather than let it lapse.
One week of paying attention, and the habit you commit to is one your actual life has room for. That is a much better starting position than willpower.