Goals versus systems: what the argument gets right and where it breaks
The case for forgetting goals and focusing on process, given fairly — followed by the part it leaves out. Where the anti-goal argument is strongest, where it is weakest, and the synthesis that actually survives contact with a real month.
One line has shaped how people talk about self-improvement over the last decade more than anything from the research literature: forget goals, focus on systems. It comes largely from Scott Adams, who framed goals as being for losers and systems for winners, and was popularised by James Clear, whose version is that you do not rise to the level of your goals, you fall to the level of your systems.
It is a very good line. It is also, in the form most people repeat it, wrong in a way that costs them something. Both sides deserve stating properly, because the useful position is neither of the ones being argued for.
The case for systems, stated as strongly as it deserves
A goal is the result you want; a system is the set of things you do repeatedly. Every runner at the start of a marathon has the same goal and they finish in wildly different times, so the goal cannot be what separates them. What separates them is what their Tuesdays looked like for the previous six months.
Three further observations, all fair:
- A goal is satisfied and then gone. Reaching a target weight and immediately reverting is nearly the default, because the goal was structured to end and the behaviour ended with it.
- A goal puts happiness in the future. Until the day you hit it you are, by your own accounting, not yet successful — a long time to spend in deficit.
- A goal you miss is a failure, whereas a process you kept is a process you kept. The same six months of running is a triumph or a disaster depending on which side of an arbitrary time your finish fell.
And there is a real point underneath it. Time spent thinking about the outcome is not time spent doing the thing, and there is a recognisable failure mode — visible in anyone who has planned a goal beautifully and executed none of it — where the outcome becomes the object of contemplation rather than the reason for action.
The counterpoint that usually gets left out
Here is the honest part. This is aphorism, not finding. "Winners and losers have the same goals" is a memorable sentence produced by a cartoonist making an argument, not a result from a study. It may well be true; it has not been shown to be true in the way a research claim is shown.
And the actual research tradition points the other way. The most robust finding in this area — Locke and Latham, replicated across a great many studies — is that specific, difficult goals produce better performance than vague intentions to do your best. "Just show up and follow the process" is, from that literature's point of view, uncomfortably close to "do your best", the condition the goal is supposed to beat.
That literature has its own limits and this page will not pretend otherwise: mostly short, well-defined tasks with clear feedback, often with the goal assigned by someone else, in laboratories and workplaces rather than across a year of ordinary life. But if you are going to appeal to evidence at all, the evidence is on the goals side, and the systems argument is appealing to something else — observation, plausibility, and an intuition many of us share.
The specific danger, stated plainly: a system with no target cannot tell you it is failing. If your system is "practise guitar for twenty minutes daily" and you do that for two years while playing the same three songs, the system reports perfect success throughout. Only a target — play a whole song from memory at a friend's party — can inform you that twenty faithful minutes a day have been going into the wrong twenty minutes. Activity for its own sake is the failure mode of the systems view, and it is quiet, which makes it worse than the failure mode of goals. A goal you miss at least announces itself.
The synthesis that actually holds
The two are not competitors but answers to different questions, and the confusion comes from arguing as though you must pick one.
The goal sets the direction, and — this is the part usually missed — it supplies the criterion by which you judge whether the system is working. The system is what your Tuesday consists of. A goal without a system is a wish with a date attached. A system without a goal is a habit with no error signal, which will run happily for years in a direction you did not choose.
This is why Northstar holds both at once rather than choosing. A goal carries a deadline and a measure, which is the direction and the criterion. Underneath it sit steps marked once, daily or weekly, which are the system — the recurring ones keep returning without being re-typed. And its seven and thirty-day review reports what share of your recorded effort was actually attached to the goal rather than to everything else the month contained. That figure is the one the systems-only view cannot produce, and it is frequently uncomfortable.
Where the anti-goal argument is strongest
It deserves the win in several places, and being specific about them beats being even-handed in general.
- All-or-nothing goals. "Run every day this year" has exactly two states and one week of flu moves you permanently into the failed one. Here the process framing is straightforwardly better: run most days, keep running.
- Goals whose outcome you do not control. Getting the promotion, winning the tender, the book selling. You control the submissions, the practice, the preparation. Attaching your sense of whether to continue to something decided by other people is a reliable way to stop.
- Identity-level changes. "Become someone who reads" has no finish line, and inventing one — fifty books — quickly produces the behaviour of finishing thin books in December.
- Open-ended crafts. Writing, drawing, an instrument, cooking. There is no state of being done, and a target mostly distorts the practice.
Where it is weakest
Equally specific, in the other direction.
- Anything with a fixed external endpoint. The exam is in May, the race is on the third, the application closes. Good process does not help if it is not calibrated to a date that exists whether or not you acknowledge it.
- Deadlines with real consequences. A thesis, a launch, a tax return. "Trust the process" is a lovely thing to say in March and an expensive thing to have said by August.
- Projects rather than habits. A single repeating routine cannot carry one, because week twenty requires an activity that did not exist in week two.
- Anything where you might be practising the wrong thing. Improvement is not automatic with time spent, and without a target that tests you, a comfortable routine can be indefinitely self-confirming.
How to actually set it up
- Write the outcome with a date and a measure. One sentence — direction and criterion, not a promise, and allowed to be wrong.
- Write the two or three recurring actions that would move it. These are the system, sized to survive your worst week rather than to impress you on your best.
- Do not track the outcome daily. Weight, revenue and totals are noisy over short intervals, and reading noise as feedback is how people abandon working systems. Track the actions daily; look at the outcome at phase boundaries.
- Set a review point. Weekly is enough. The question is not "am I on target" but "did my days go where my goal was".
- When they diverge, change something on purpose. Either the system is not big enough to reach the goal, or the goal is not one your current life can produce. Both are legitimate findings, and choosing between them consciously is the entire value of having both halves.
What neither of them fixes
A goal you do not actually want will not be rescued by a system, and a system for something you do not care about will be abandoned however well it is designed. A lot of the energy people put into this argument is displacement activity around a goal they inherited from someone else and have not yet admitted they can drop.
And neither survives circumstances. Illness, a newborn, a punishing quarter at work break plans and processes alike, and the correct response is to shrink whatever you have until it fits the life you currently have. In those months the question of which framing is superior is beside the point.
Frequently asked questions
Are systems really better than goals?
Neither is better; they answer different questions. The goal supplies direction and the criterion for judging whether the process is working. The system is what your days actually consist of. A system with no target cannot tell you it is failing, which is the quiet failure mode people miss.
Is the "systems beat goals" idea backed by research?
Not directly. It comes from Scott Adams and was popularised by James Clear as an argument and an observation. The formal research tradition — goal-setting theory — points the other way, finding that specific, difficult goals beat vague intentions, though mostly in short, well-defined tasks rather than across a year of ordinary life.
When should I drop the goal and just follow a process?
When the outcome is not under your control, when the goal is all-or-nothing so one bad week kills it, or when the pursuit is open-ended — an instrument, writing, reading more. Keep the goal when there is a fixed external date or a real consequence attached to missing it.
How do I know if my system is working?
Compare it against something. Track the actions daily and the outcome only at phase boundaries, since outcomes are noisy over short intervals. Then ask, weekly, whether your recorded days went where your stated goal was — a question that needs both a record and a target to answer.
Try it on a real goal
Northstar cuts a goal into phases and phases into steps you write yourself, then names one next action a day. Its statistics come from your own entries, and it shows an empty state rather than a placeholder number. Everything stays on the device — no account, no sync, no analytics, no notifications. Free, with no in-app purchases.