Goal setting: most people write a schedule instead

Half of what gets written down as a goal is a schedule wearing a goal's clothes. That single confusion accounts for a lot of abandoned Januaries.

Goal-setting advice is mostly about phrasing. Make it specific, make it measurable, put a date on it. That is worth doing and it takes about ten minutes, which tells you something: if phrasing were the constraint, everybody who has heard of SMART goals would be finished by now.

The gap is between the goal and the next ordinary Tuesday. A goal describes a destination. It does not tell you what to do at 7am when you are tired, and it does not notice when you skip.

A goal is an outcome, not a schedule

Most goal-setting goes wrong before the framework is applied, because the thing being written down is not a goal. "Run three times a week" is a schedule. "Read twenty pages a day" is a schedule. Both are things you do repeatedly and neither of them names a result.

A goal is the outcome you would be pleased about: run a 10k without stopping, hold a conversation in Spanish, ship the side project. It happens once. The schedule is how you get there, and confusing the two is expensive, because a schedule with no outcome behind it has no way of telling you whether three runs a week is enough, too little, or the wrong exercise entirely.

What SMART gives you, and what it leaves out

SMART is a checklist for the sentence. Specific, measurable, achievable, relevant, time-bound. Run "get fit" through it and you get "run a 10k without stopping by the end of March", which is a real improvement: it names an outcome and a date, and you will know on the 31st whether it happened.

What it does not give you:

• the first action, small enough that you will actually do it this week
• a checkpoint before the deadline, so a bad month is visible in week two rather than week ten
• anyone who notices either way

Those three are the difference between a goal you achieve and a goal you rewrite next January with a different verb.

Where SMART came from

SMART was published in November 1981, in Management Review, in a short piece by George T. Doran called "There's a S.M.A.R.T. way to write management's goals and objectives". It was advice for managers writing objectives. Two of the letters did not mean what they mean now.

• Specific, Measurable, Time-related. Target one area, quantify progress or at least suggest an indicator of it, put a timeline on it. These three came through intact.
• Assignable. Say who does it. This letter disappeared, because a person setting a goal for themselves already knows who does it. Achievable moved into the slot.
• Realistic. Attainable with the resources you have. Now usually read as relevant, a different question.

Doran also noted that objectives cannot always be quantified at every level of an organisation, and argued for balancing the countable ones against more abstract goals. That caveat fell off in transit. The acronym survived.

In 2023 Swann and colleagues reviewed SMART as it is used in physical activity promotion and listed the problems: not based on scientific theory, not consistent with empirical evidence, does not consider what type of goal is set, not applied consistently, redundant in its criteria, and not being used as originally intended. None of that makes SMART worthless. It makes it what it started as, a house style for writing an objective down, invented for a manager with a filing cabinet. The mistake is promoting it to a method.

The letter that argues with the evidence

Achievable, the letter that replaced Assignable, is the weakest of the modern five and it points the wrong way. Edwin Locke and Gary Latham spent decades on the question and summarised the results in American Psychologist in 2002. Across their meta-analyses, the relationship between goal difficulty and performance is positive and linear: the hardest goals produced the highest effort and the highest performance, with effect sizes from 0.52 to 0.82. Performance levelled off or fell at two points only. When people reached the limits of their ability, and when commitment to a very hard goal lapsed.

So "keep it achievable", read the way most people read it, has a price. Pick the target you are already confident about and you get the performance of a target you were already confident about. Realistic was doing narrower work: do not set a goal beyond your ceiling, and do not set one you will stop caring about by week three. Difficulty belongs in the outcome. Keep it out of what you have to do on a Tuesday.

What the research supports

The same paper compared specific, difficult goals against the instruction most people give themselves, some version of "do your best".

Specific and difficult won consistently, with effect sizes from 0.42 to 0.80. Locke and Latham explain it in one line: when people are asked to do their best, they do not do so. A do-your-best goal has no external referent, so everyone defines it privately and a wide band of performance stays acceptable. Naming a level closes the band.

The scale is worth knowing, because most goal advice has none behind it: well over 100 different tasks, more than 40,000 participants, at least eight countries, time spans from one minute to twenty-five years. Two conditions decide whether any of it reaches you.

• Commitment. The link between goal and performance is strongest when you are committed, and commitment counts for most when the goal is hard. Making it public raises it.
• Feedback. You need summary feedback that shows progress against the goal. Goals plus feedback beat goals alone. Without it you cannot tell whether to push harder or change approach, so effort goes into the same failing plan.

Those two are why the rest of this page is about checkpoints and other people, not phrasing. A milestone with a date is feedback. A crew that can see whether today is logged is the public half.

Split it until one piece fits in a day

A goal you can do nothing about today is not yet a plan. Keep splitting until one piece fits into a normal weekday, then look only at that piece.

"Run a 10k without stopping by the end of March" splits into three layers:

• The goal. 10k, unbroken, by 31 March. One event, one date.
• Milestones. 5k without walking by mid-January. 8k by the end of February. Each is a checkpoint with a date, early enough that failing one still leaves time.
• The habit. Three runs a week, twenty minutes to start. This is the only layer that appears on today's list.

The 10k never shows up in your day. That is the point: you are not supposed to be thinking about it on a Tuesday, you are supposed to be going for a twenty-minute run.

CommitMe does this splitting for you. You write the goal badly, in one sentence, and it comes back as three to five milestones and one habit to do this week. You can rewrite any of it. Starting from something concrete beats starting from a blank page.

When a hard, specific goal makes you worse

There is a real exception, and summaries drop it. Goal effects shrink as the task gets more complex. In the meta-analyses the specific-difficult advantage was 0.77 on the simplest tasks and 0.41 on the most complex.

It can do more than shrink. Kanfer and Ackerman ran an air traffic controller simulation in 1989 and found a performance target interfered with acquiring the knowledge the task needed. The group told to do their best did better. Winters and Latham later showed the fault lay with the type of goal, not with goal setting: set a hard, specific learning goal instead of a performance one and the usual result comes back.

The rule you can carry: if you do not yet know how it is done, do not set a number. "8k without walking by the end of February" works when you know the territory, and a hard one works better. When you do not, make the first milestone "have a training plan I understand and have run for two weeks", and let the number wait until you have it.

CommitMe writes the first milestones for you and every one is editable. If the first comes back as a number and you have never done this before, replace it with something you can find out.

How many goals: fewer than you have written down

The most common mistake after confusing a goal with a schedule is running five at once. Each one looks reasonable in isolation. Together they compete for the same evenings, and the competition is settled by whichever felt most urgent that day, so all five advance slightly and none finishes.

One or two at a time, with the rest written down somewhere you can see them but not working on them. A list of ten goals is not ambition, it is an undeclared decision to fail at eight of them.

Goals have side effects

In 2009 four researchers published a paper in Academy of Management Perspectives arguing that the harms of goal setting had gone unexamined. They called it "Goals Gone Wild". The side effects they name: a narrow focus that neglects everything outside the goal, a rise in unethical behaviour, distorted risk preferences, corrosion of culture, and reduced intrinsic motivation. Their recommendation is that goals be handled as prescription-strength rather than over the counter. Their examples are organisational. Sears put a quota of $147 an hour on its auto repair staff, and staff overcharged customers and carried out repairs nobody needed. The chairman later accepted the goal had produced it.

The personal version is quieter and you have probably run it. You set "read 30 books this year" and start choosing short books. The number gets hit and the reason you picked it does not. Keep the goal as close as you can to the outcome you actually wanted and not to a proxy for it, and take it seriously the first time hitting the number stops feeling like the point.

Decide now what counts as done

Write the finish line before you start, in terms someone else could adjudicate. This is the difference between a goal and a wish, and it is where most goals quietly rot: not abandoned, just left vague enough that you never have to admit either way.

• Bad: "get fitter." No finish line, so it can absorb any amount of effort and never be reached.
• Better: "run 10k." Still ambiguous. On a treadmill? With walking breaks?
• Good: "run 10k outdoors without stopping to walk." A stranger could watch and tell you whether it happened.

Do the same for each milestone. In CommitMe a milestone is reached or not reached, with no partial credit, because a checkpoint you can argue with is not a checkpoint.

What to do when you miss the date

You will miss some. The useful question is which of three things happened, because they need different responses.

• The habit did not happen. The plan was fine and the execution was not. Keep the goal, shrink the daily action until it survives a bad week.
• The habit happened and the milestone still slipped. The plan was wrong. Three runs a week was never going to produce that time. Change the plan, not your opinion of yourself.
• You stopped caring. The most common one and the least discussed. Drop it properly, in writing, rather than letting it sit on the list generating guilt for another year.

Then change what is around you

A goal that depends on deciding correctly every evening is fragile no matter how well it is written. The reliable move is to alter your surroundings so the default does the work, which is a subject of its own, covered in self discipline and what actually works.

Checkpoints beat deadlines

A deadline in March tells you nothing in January. Milestones are checkpoints: each one is a date where you either passed or you did not, early enough that the answer still changes what you do. Miss the first and you have ten weeks to adjust. Miss the deadline and you have a story about why this year was unusual.

The written-goals study that does not exist

You will meet this one eventually. In 1953 Yale surveyed its graduating class, 3% had written goals, and twenty years later that 3% was worth more than the other 97% combined. There is a Harvard MBA version with the same shape and a factor of ten in it.

It never happened. Fast Company went looking for it in 1996 and came back empty. Nobody since has produced a questionnaire, a cohort or a paper.

The best witness is Gail Matthews, a psychologist at Dominican University of California, because hers is the research now cited in its place. She opens her own abstract by calling the study an urban myth, and reports that reviews of the literature by her and by the Harvard social psychologist Steven Kraus, alongside the Fast Company reporting, found no such study. The myth is the reason she ran hers. The laundering carries on regardless: the longest article you will find on the psychology of goal setting still describes her work as a search for evidence supporting claims out of Harvard Business School. That is the reverse of what she wrote.

Her study deserves accurate handling too. 267 participants across five conditions, running from goals kept in your head up to goals written down with action commitments and weekly progress reports sent to a friend. The last group did best. It was presented at the Western Psychological Association convention in 2007, so it is a conference talk rather than a journal paper. It points the same way as the older and much larger commitment literature. Writing it down does something and telling someone does more. That is a direction, not a statistic, and a direction is all a study that size can give you.

The part nobody writes down

Every goal framework assumes you will do the thing once it is written clearly. That assumption is doing most of the work, and it is the assumption that fails. What changes the odds is someone else knowing what you said you would do.

In CommitMe a habit you share is visible to a crew you invited, up to five people, and they see whether today is logged. Nobody watches you run. They see a green square or they do not, and the prospect of being the person who did not log it is doing the work that a well-phrased sentence cannot.

When you miss

You will miss. A framework that treats a missed week as failure is one you abandon in February, roughly when most goals die. CommitMe gives you a pause day for every four completions, banked up to thirty. Opening an absence spends them and the streak carries through, so a week of flu costs you a few banked days rather than the whole record.

Start free