Every goal points in one of two directions: toward something you want, or away from something you fear. That direction shapes how it feels to pursue, how well it works, and how easily it tips into doing harm.
The most basic fact about goals is that they have a direction. Some pull you toward a reward, finish the marathon, win the account, learn the language. Others push you away from a threat, do not get fired, do not embarrass yourself, do not fall ill. This is the approach versus avoidance distinction, and it runs deep: at the level of self-regulation, behaviour works like a set of feedback loops, approach loops shrink the distance to a desired end, avoidance loops widen the distance from a feared one (Carver and Scheier, 1998).
The same split shows up in how people are wired to chase things. Tory Higgins called it regulatory focus: some people, and some situations, lean toward promotion, framing life in terms of hopes, gains, and what could be won, and pursue goals with eagerness, while others lean toward prevention, framing life in terms of duties, safety, and what could be lost, and pursue goals with vigilance (Higgins, 1997). Neither is a defect; they are two different operating modes.
It is tempting to conclude that 'approach' is simply the better mode, and there is real evidence pointing that way: goals framed as avoiding failure tend to drain both enjoyment and performance, while goals framed as approaching mastery tend to lift them (Elliot and Church, 1997). But the deeper finding is about fit, not direction. When the way you pursue a goal matches your focus, eager, gain-framed means for a promotion mood, careful, safety-framed means for a prevention mood, the pursuit feels more right and you value the outcome more (Higgins, 2000). A prevention focus is exactly what you want in a surgeon or an auditor. The cost comes from a mismatch, not from avoidance as such.
Knowing a goal's direction is one thing; getting from the goal to the act is another, and here three tools stand out for how reliably they work. The first is plain goal-setting: decades of studies show that specific, challenging goals produce more than vague "do your best" ones, because they direct attention and effort and make progress measurable (Locke and Latham, 2002). The second is probably the most useful trick in this whole article. People who merely intend to act often do not, but those who form an implementation intention, a concrete if-then plan ("when I get home from work, I will put on my running shoes"), follow through far more often. Across hundreds of studies the effect is medium-to-large (Gollwitzer, 1999; Gollwitzer and Sheeran, 2006), because the plan hands control of the action to a situation rather than to willpower, creating something close to an instant habit. The third is identity: a goal that feels like an expression of who you are pulls harder, and, in a neat twist, when an action feels identity-congruent, hitting difficulty reads as a sign that it matters ("this is hard because it is important, and it is for people like me") rather than as a signal to quit (Oyserman and Destin, 2010).
The cleanest warning in the goals literature is that the most celebrated tool, the specific hard target, has a serious dark side. In a paper pointedly titled "Goals Gone Wild", researchers catalogued the systematic damage that aggressive goal-setting can do: a narrow focus that starves everything not being measured, a rise in unethical behaviour and outright cheating to hit the number, distorted risk-taking, eroded teamwork and culture, and the slow draining of intrinsic motivation (Ordóñez et al., 2009). Their image is that goal-setting is not a harmless vitamin but a prescription-strength drug that needs careful dosing and a warning label. The defenders of goal-setting argue the dangers are manageable with good design, and the debate is genuinely live, but the practical lesson is settled enough: a target changes behaviour, including the behaviour you did not ask for. This is the everyday face of Goodhart's law, when a measure becomes a target, people optimise the measure rather than the thing it was standing in for.
The second qualification is the one already flagged: approach is not automatically better than avoidance. The headline that avoidance goals are bad is true on average and misleading as a rule, because a prevention focus is the right tool whenever the job is to not make mistakes, and the value of any goal pursuit depends on the fit between the goal and the person chasing it.
There is also the familiar gap between a goal and the deed. Stated goals predict behaviour only modestly, which is the whole reason implementation intentions matter, and most of the evidence here comes from self-report and shorter-term studies, with the identity-based strand still younger and thinner than the rest.
Much of the tool evidence is from Western samples and relatively short horizons, and the link from a goal to a sustained change of behaviour is harder than a single study suggests. The "set a specific hard goal" advice in particular travels badly from simple, well-defined tasks (where it shines) to complex, learning-heavy ones (where it can backfire), so the rule needs the context attached.
What is the right dose of a hard goal before its side effects outweigh its benefits? How durable are identity-linked goals once the identity stops being salient? And when is a prevention, avoidance-shaped goal genuinely the better choice rather than a handicap to be reframed away?
The usable core: read a goal's direction and match how it is pursued to the person, use the proven tools (specific, planned, identity-linked) to cross the gap from intention to action, and treat aggressive targets as powerful and dangerous rather than simply good.
Two levers and one warning. Match the framing to the audience and the product: promotion-framed messaging (what you will gain, who you could become) lands with a promotion-focused buyer, while prevention-framed messaging (protect what you have, do not miss out, stay safe) lands with a prevention-focused one, and that fit measurably raises both persuasion and willingness to pay. To move behaviour rather than just intention, build implementation-intention prompts into the journey ("the next time you open the app, set your delivery day") instead of relying on the customer's resolve. The warning is about internal targets: specific, stretch goals drive performance and, past a point, drive gaming, corner-cutting, and tunnel vision (Ordóñez et al., 2009), so pair every hard target with guardrails, ethics checks, and a few non-target measures, or you will hit the number and lose the plot.
Promotion and prevention are the two great framings of any campaign: a politics of hope and gain ("a better future we can build") versus a politics of protection and risk ("defend what we have, do not let them take it"), and each resonates with a different mood in the electorate. Identity-linked goals mobilise far better than abstract appeals, since "people like us vote, and this is what we stand for" turns a chore into an expression of self. And in office, the goals-gone-wild lesson is a warning about governing by target: crude numerical goals get gamed by the very services they were meant to improve.
The behaviour-change toolkit here is unusually practical: implementation-intention prompts measurably lift everything from vaccination to voting to bill payment, and they cost almost nothing to add to a letter or a form ("write down when and where you will go"). Identity framing and a single specific ask help too. But the central caution belongs to the state more than anyone, because government runs on targets, and the literature is a direct warning that waiting-time targets, test-score targets, and quota systems reliably produce the gaming, narrowing, and distortion of "goals gone wild" unless they are designed with the side effects in mind.
Three habits. First, notice a goal's direction and match the pursuit to the person, eager means for the hopeful, careful means for the cautious, rather than assuming approach is always right. Second, do not stop at intention: make the goal specific, attach an if-then plan that names the trigger, and tie it to an identity. Third, respect the dark side of hard targets: the more aggressive the number, the more you need guardrails against the behaviour you did not intend to reward.
.png)