How to Set Goals You Will Actually Stick To

Direct answer: Goals stick when they are specific and moderately difficult, per Locke and Latham’s goal-setting theory, one of psychology’s most replicated frameworks. Vague goals (“do your best”) and overly easy ones underperform clear, challenging targets. Difficulty must pair with real commitment and feedback; beyond a person’s ability, harder goals stop helping and start hurting performance.

What Locke and Latham’s Research Actually Found

Edwin Locke and Gary Latham spent over three decades testing how goal design affects performance, synthesized in their 2002 American Psychologist paper covering hundreds of studies. Their central, consistently replicated finding is that specific, difficult goals produce higher performance than vague goals or the common exhortation to “do your best.” Meta-analytic effect sizes for goal difficulty range from d = .52 to .82, a large effect by psychology standards. That gap matters for a family setting a goal together: “get better at math” gives a child nothing to aim at, while “finish twenty practice problems by Sunday” gives a target that can be hit, missed, or tracked.

The relationship between difficulty and performance is roughly linear up to the point where a goal exceeds a person’s ability or the necessary commitment isn’t there. Past that point, harder goals stop helping. This is why doubling a goal’s difficulty is not automatically better; there is a ceiling set by what someone can actually do and how much they genuinely care about doing it. Locke and Latham’s theory also names the conditions that make difficulty productive rather than demoralizing: the person must be committed to the goal, receive feedback on progress, and have sufficient ability and low task complexity for the goal to translate into a real strategy. A child who wants the goal, gets regular feedback on how they are doing, and is not facing a task far beyond their current skill level is in the conditions where a difficult goal helps. Remove any one of those conditions and the same goal can become discouraging instead of motivating.

It is worth being precise about what this theory covers and what it does not. This is a theory about how to word and size a goal, not about the follow-through habits, reminder systems, or environmental design that make people act on a goal day to day. A perfectly specific, perfectly calibrated goal still needs some mechanism for follow-through; goal-setting theory explains why the target itself works, not how to build the daily behavior around it.

The 21-Day Myth vs. What the Data Shows

The claim that habits form in 21 days traces back to Maxwell Maltz’s 1960 book Psycho-Cybernetics, where he observed that his cosmetic surgery patients, and amputees adjusting to phantom limb sensations, took a “minimum of about 21 days” to adjust. Maltz’s own word was minimum, describing a floor, not an average. The self-help industry stripped that nuance and turned it into a fixed universal rule that gets repeated as fact in countless goal-setting guides.

Real data tells a different story. A 2010 University College London study by Phillippa Lally and colleagues tracked 96 volunteers forming a new daily habit over 12 weeks and found that automaticity built along a gradual, decelerating curve, not a switch that flips on day 21. The average time to reach near-peak automaticity was 66 days, and the range across individuals and behaviors was 18 to 254 days. That range is the part most retellings of this study leave out: some people and some behaviors settle in within a few weeks, while others take the better part of a year, and both outcomes are normal rather than a sign of failure.

For a family setting a goal, the practical takeaway is about expectations rather than technique. A goal that quietly assumes “21 days and it’s automatic” sets people up to read normal, slower progress as failure, and a child who is still finding a new routine effortful on day 30 has not necessarily done anything wrong. Sizing expectations around a wide, individual range rather than a fixed three-week deadline keeps a stalled morning or a skipped day from feeling like the goal has collapsed.

Why “1% Better Every Day” Oversells the Math

The viral claim that improving 1% daily compounds into being 37 times better in a year comes from the arithmetic of 1.01 raised to the 365th power, popularized in James Clear’s Atomic Habits. The multiplication itself is correct as arithmetic, but it is a mathematical illustration, not a peer-reviewed finding about how real skill or behavior change actually unfolds. No study shows a fixed 1% daily gain compounding smoothly for a year in any real domain, whether that is reading speed, a musical instrument, or a fitness goal.

Decades of skill-acquisition research instead describe a decelerating curve: performance improves quickly early on, then flattens as gains get harder to find. This pattern was formalized as the power law of practice by Allen Newell and Paul Rosenbloom in 1981 and has been revisited by later researchers who confirmed the diminishing-returns shape even while debating its exact mathematical form. In plain terms, the first few weeks of a new goal, whether it is a child learning a musical instrument or a parent building a new exercise routine, tend to produce visible, encouraging jumps, and then progress slows even though effort stays the same. That slowdown is not evidence the goal has stopped working; it is closer to the expected shape of how skill actually develops.

This lines up with Locke and Latham’s own guidance rather than contradicting it. A goal should be pitched at the edge of real, current ability, difficult but attainable, and reassessed as ability changes, not benchmarked against a tidy exponential curve that no one’s actual trajectory follows. A family goal that expects steady, compounding daily improvement forever is setting a target the underlying skill-acquisition research does not support; a goal that expects fast early gains followed by a longer, flatter stretch of steady practice is closer to what the evidence actually shows.

Verified goal-setting research against these sources on 2026-07-31: Locke, E.A. & Latham, G.P. (2002), “Building a Practically Useful Theory of Goal Setting and Task Motivation,” American Psychologist, 57(9), 705-717 (https://pubmed.ncbi.nlm.nih.gov/12237980/), supporting the finding that specific, difficult goals outperform vague goals and ‘do your best’ instructions, with meta-analytic goal-difficulty effect sizes of d=.52 to .82. Lally, P., van Jaarsveld, C.H.M., Potts, H.W.W. & Wardle, J. (2010), “How are habits formed: Modelling habit formation in the real world,” European Journal of Social Psychology, 40(6), 998-1009 (https://onlinelibrary.wiley.com/doi/abs/10.1002/ejsp.674), supporting the finding that habit automaticity forms on an asymptotic curve averaging 66 days, ranging from 18 to 254 days across individuals and behaviors, directly debunking the 21-day myth. University of Surrey news interview with study author Dr. Pippa Lally (https://www.surrey.ac.uk/news/does-it-really-take-66-days-form-habit-we-asked-expert-dr-pippa-lally), corroborating the 66-day average and wide individual range directly from the researcher and explaining why the 21-day figure is a myth. Newell, A. & Rosenbloom, P.S. (1981), “Mechanisms of Skill Acquisition and the Law of Practice” (https://www.researchgate.net/publication/243783833_Mechanisms_of_skill_acquisition_and_the_law_of_practice), supporting the finding that skill and performance improvement follows a decelerating, diminishing-returns curve rather than smooth exponential compounding, countering the literal ‘1% daily equals 37x per year’ claim.