You do not need January. Fresh-start research shows goal behaviour spikes after temporal landmarks of every size — a new month, a new week, a birthday — so the nearest Monday works as a line in the sand. Audit what actually happened in the last six months using data rather than memory, cut the goals that were never really yours, and reframe the survivors as one 90-day project with a start, a finish and a written miss policy.
There is a specific flavour of guilt that arrives in late summer: the goals you set in January are still technically alive, in the sense that nobody has officially declared them dead. The gym membership renews. The book sits at page 40. And there is a comforting voice suggesting that with four months to go it is a bit late to start, and January is not so far away really. I ran that exact loop for most of my thirties, and it cost me years — not because I lacked plans, but because I kept scheduling the restart for a date the calendar had to approve.
Is it worth restarting goals in the middle of the year?
Yes, and the evidence is more specific than motivational advice usually is. Field research on the fresh-start effect found that goal behaviours — gym visits, diet searches, commitment-contract signups — spike after temporal landmarks: new years, but also new months, new weeks and birthdays. The mechanism is that a landmark separates you from the version of yourself who failed, and any landmark can do that job. January is simply the biggest one, not a uniquely effective one. The next Monday is a landmark. So is the first of next month. So is the day you finish this article.
Why did the January version die?
Usually not for the reason you have been telling yourself. Annual goals fail in a characteristic way: they are set at a scale with no deadline pressure until the deadline is unreachable, they carry no review cadence so nobody notices the drift for months, and they were often chosen in a burst of January optimism by a version of you with different constraints. Then one bad week arrives, the restart ritual demands a clean date, and the goal enters the long undead phase where it is neither pursued nor abandoned. That undead phase is the actual problem. It costs attention all year while producing nothing.
How do you audit six months honestly?
With data, not memory, because memory will either flatter you or flatten you and both are useless. The sources are already on your phone: step counts and workouts, screen-time reports, bank transactions, photo timestamps, reading history, the calendar itself. Give it an hour, pull the numbers for January to now, and write one honest line per goal — what actually happened, not what you intended. Only then decide. About half of everyday behaviour runs on defaults keyed to context rather than decisions, which means an audit is not a character assessment; it is a wiring diagram of the life you have been living by accident.
Estimates across the paper’s methods ranged ~35–45% — “about 45%” is how the authors themselves cite it later.
| What the audit shows | What it usually means | The move |
|---|---|---|
| Started strong, dead by March | The goal was real, the frame was not — no finish line, no miss policy | Keep it; re-launch as a single 90-day project with both |
| Never actually started | It was someone else’s goal, or an identity you liked more than the work | Cut it. Writing it down again in January will not change this |
| Ran quietly all year, unmeasured | It already fits your defaults — you just have no evidence of it | Keep and instrument it; you may be further along than you think |
| On and off, never fully failed | A scheduling mismatch, not a motivation problem | Re-anchor it to a cue you already have, and cut the target in half |
| Replaced by something you did not plan | Your actual priorities moved and the list did not | Promote the real one onto the list and drop the ghost |
Why a 90-day frame beats what is left of the year?
Because four months is long enough to hide in and a quarter is not. A 90-day frame gives you a start date, a finish date and a review you cannot postpone — and it comfortably covers the timeline habits actually run on: in the best real-world study, the median time to automaticity was 66 days, with a range of 18 to 254. So one quarter gets the median habit installed with margin to spare. Pick one goal per life area, three at most, with the smallest version of each that would still count. Then write the miss policy before day one, because the policy you agree to in advance is the only one that survives a bad week.
And put a monthly review in the calendar as an event, not an intention. Fifteen minutes, three questions: what does the data say, what is quietly dead, and what is the one thing for next month. The review is what keeps a goal from entering the undead phase — most goals do not fail on a specific day, they fail by never being looked at again.
The mid-year honesty check
- I could not say, without looking, what my January goals actually were
- At least one goal on my list has had zero activity since spring
- I have been waiting for a clean date — a Monday, a month, next January — to restart
- None of my goals has a written finish date
- I have never reviewed them in a scheduled sitting
- I have no data for the last six months beyond how I feel about it
Interactive tool — check items off and get a verdict (loads with the app version of this page).
Quick answers
Is it too late to start goals in August?
No — and waiting for January is the more expensive option, because it costs you four months to buy a slightly better starting feeling. Fresh-start research found that goal behaviours spike after temporal landmarks of every size: new years, new months, new weeks, birthdays. The psychological work a landmark does is to separate you from the version of yourself who failed, and a Monday does that adequately. Practically, four months is more than enough for a full 90-day block plus a review — and the median habit takes about 66 days to become automatic, so a habit started this month is an installed habit by the end of the year rather than a resolution waiting for one.
How do I restart a goal I have already failed at this year?
Change three things before you touch the goal itself. First, give it a frame: a start date, a finish date roughly 90 days out, and the smallest version that would still count — most abandoned goals were never sized to a real week. Second, write the miss policy in advance, in words: what happens when you miss one day, and what happens when you miss three. The failure is almost never the miss, it is the improvised restart ritual that follows it. Third, anchor it to a cue you already have rather than a time you hope to defend. And log it somewhere visible, because a goal with no record cannot show you the difference between week three and week zero.
How many goals should I set for the rest of the year?
Three, maximum, and one is a perfectly respectable answer. The reason is arithmetic rather than philosophy: a new habit consumes attention and self-control until it becomes automatic, and the median in the best real-world study was 66 days, with a range up to 254. Six simultaneous goals means six young habits competing for one adult’s attention during exactly the period when each is most fragile — which is why so many lists collapse as a stack by the second Thursday. Pick one goal per life area at most, choose the smallest version that would still count, and add the next only after the first has survived a genuinely bad week.
What is a mid-year review and how do I actually do one?
It is one scheduled hour where you replace your memory of the last six months with your data. Pull what you already have — step counts, workouts, screen-time reports, bank transactions, photo timestamps, your calendar — and write one honest line per goal about what actually happened, not what you intended. Then sort each goal into keep, cut or reframe, and be generous with cut: a goal that never started in six months is usually someone else’s goal, and writing it down again in January will not fix that. Finish by scheduling the next review as a calendar event. Goals rarely die on a specific day; they die by never being looked at again.
Do not wait for the calendar to grant permission. Audit six months with data, cut what was never yours, and relaunch the survivors as one 90-day project with a start, a finish and a written miss policy — then put the review in the calendar so the next one cannot quietly die.
Ready to stop reading and start measuring?
Everything on this blog is one tap away in the app — the forgiving streak, the 67-day protocol, the proof.