Why most goals should be smaller

Goals are set at a moment of enthusiasm, sized to that enthusiasm, and then pursued at ordinary energy levels for months. The mismatch is predictable and rarely planned for.

The planning fallacy

One of the most robust findings in judgement research: people systematically underestimate how long tasks will take and how much they will cost, even when they have relevant past experience.

The effect persists when people are told about it, which is unusual and slightly humbling.

The proposed mechanism is that planning is done by imagining the task going well — the inside view — rather than by referring to how similar tasks have actually gone.

The correction that works is the outside view: asking how long this took last time, or how long it takes people generally, rather than how long it should take.

That single substitution improves estimates substantially and almost nobody does it.

Why a smaller goal outperforms

It survives a bad week. The decisive property.

A plan requiring five sessions a week fails in the week containing illness, travel or a crisis. A plan requiring two does not, and the two-session plan executed for a year beats the five-session plan executed for six weeks by a wide margin.

It preserves the cue. Consistency builds automaticity; intensity does not. A behaviour maintained at low volume remains a habit, and habits are what carry effort through periods of low motivation.

It avoids the collapse. Missing an ambitious target repeatedly produces the conclusion that the goal is unachievable, which produces abandonment.

Hitting a modest target repeatedly produces the opposite, and — per the self-efficacy literature — the accumulated evidence of success is what makes larger targets feel possible later.

The objection

Worth taking seriously: the goal-setting research finds that difficult goals produce better performance than easy ones.

That finding is real and its conditions matter. It applies to defined tasks, within a person's ability, with feedback, over relatively short horizons, where commitment is present.

A personal goal pursued over a year, with no external accountability, through varying life circumstances, is a different situation from a work task with a deadline and a manager.

For those conditions, consistency dominates intensity, and the binding constraint is not effort per session but whether the sessions continue to happen.

Sizing it properly

A useful heuristic: choose the version you could do on your worst realistic week.

Not your average week. The week with a deadline, a cold and a family obligation.

That version is the commitment. Anything above it on a good week is a bonus rather than the standard, which means good weeks feel like success rather than like the baseline you failed to meet in the following bad one.

Minimums and maximums

A structure that works well.

Set a floor — the amount you do regardless — and a ceiling, which prevents an enthusiastic period producing an unsustainable pace and the crash that follows.

The ceiling is the part people resist and it is what stops the boom-and-bust cycle that characterises most abandoned plans.

Scaling up

Once the behaviour is automatic — which takes considerably longer than people expect, as discussed elsewhere — increasing it is cheap, because the cue and the routine already exist.

Increasing before that point is expensive, because it puts load on a link that is not yet holding.

Which produces a sequence: establish small, wait until it is automatic, then increase. Most people invert it, starting large and reducing under pressure, which never establishes the link at all.

The uncomfortable arithmetic

Two sessions a week for a year is a hundred sessions. Six sessions a week for six weeks is thirty-six.

The second feels far more serious at the point of setting, and the first is the one that changes anything.

The size that is too small

The counterweight, because the argument can be taken too far.

A goal small enough to be trivially met stops functioning as a goal. It produces no sense of progress, and it will not survive competition from anything more interesting.

The workable range is something that is genuinely uncertain but plausible — difficult enough to require attention, near enough that failing is not the default expectation.

General information about behaviour, not psychological advice. Consult a qualified professional about your own circumstances.