Specific, Difficult Goals Beat "Do Your Best" Almost Every Time

The short answer

"Do your best" is the most reasonable sounding instruction in management, and one of the least effective.

Across three and a half decades of laboratory and field research, Locke and Latham found that specific, difficult goals produce higher performance than vague or easy goals, or than urging people to do their best. The relationship between goal difficulty and performance is close to linear, right up to the limit of ability.

The reason "do your best" fails is not that people are dishonest. It is that the instruction has no reference point. Any level of effort satisfies it.

What "do your best" actually does

A goal works by supplying a standard. The standard drives four things:

1. Direction. Attention moves toward goal relevant activity and away from everything else.
2. Effort. Harder goals pull more effort, in proportion to difficulty.
3. Persistence. A named standard keeps you working when you would otherwise stop.
4. Strategy. Difficult goals force you to search for better methods, because the current method visibly will not get there.

"Do your best" supplies none of these. It is idiosyncratically defined, which means it is defined downward on hard days and upward on good ones. There is no discrepancy to detect, so there is nothing to close.

What Locke and Latham established

Their 2002 paper in American Psychologist summarised 35 years of programmatic research into a usable theory. The headline finding is that specific difficult goals beat vague or easy ones. The more useful part is the set of conditions under which that holds.

Goal commitment. Goals only work if you are committed to them. Commitment rises with importance (why this goal matters) and self efficacy (whether you believe you can do it). A difficult goal you do not believe in produces worse results than an easy one you do.

Feedback. Goals require feedback on progress. Without it you cannot adjust effort or strategy. Goals and feedback are complements; neither works well alone.

Task complexity. On complex tasks, outcome goals can backfire early, because attention goes to the outcome before the method exists. On complex work, learning goals ("discover three approaches") often outperform performance goals in the early phase.

Goal conflict. Competing goals degrade performance on both. This is the failure mode of setting five ambitious targets simultaneously.

What this does not mean

It does not mean maximally hard goals are best. Difficulty raises performance up to the ceiling of ability. Beyond that, performance flattens and commitment falls. A goal you privately believe is impossible is not a stretch goal; it is a demotivator.

It does not mean specificity alone is enough. A specific easy goal caps your performance at easy. Specificity plus difficulty is the combination that matters.

It does not mean outcome goals are always right. On genuinely novel work, forcing an outcome number too early can hurt.

How to set a goal that meets the conditions

Run this checklist before committing:

• Is the target specific enough that an outsider could judge whether you hit it?
• Is it difficult enough that your current default behaviour will not get there?
• Do you actually believe it is achievable? If not, lower it until you do.
• Is there a feedback mechanism that tells you weekly whether you are on pace?
• Does it conflict with another goal you are also committed to? If so, pick one.

Why UpTicker asks you to name a 90 day target

During onboarding, UpTicker asks you to commit to a specific outcome with a date attached rather than selecting a general area of improvement. That is not a UX flourish. Every downstream feature depends on it.

A named target gives the system a reference point it can measure against. It is what lets the daily plan rank tasks by contribution to the goal rather than by urgency, which is the usual and much worse default. It is what makes the daily score meaningful, because a score without a target is a number in a vacuum. And it is what makes trade offs decidable: when two tasks compete, the one that moves the named target wins.

The date matters as much as the number. A goal without a deadline is not difficult, because it can always be satisfied later.

Frequently asked questions

Why does "do your best" not work? It has no reference standard, so any level of performance satisfies it. Locke and Latham found specific, difficult goals outperform it across decades of studies.

How difficult should a goal be? Difficult enough that your current habits will not reach it, but not so difficult that you privately believe it is impossible. Commitment collapses when self efficacy does, and commitment is a required condition.

Do I need feedback for goal setting to work? Yes. Goals and progress feedback are complements. Without feedback you cannot tell whether to change effort or change method.

What if I have several important goals? Goal conflict degrades performance on all of them. Sequence them rather than running them in parallel, or accept slower progress on each.

Why 90 days rather than a year? Ninety days is long enough for a difficult target to require real change and short enough that the deadline stays psychologically live. Annual goals tend to lose their deadline pressure for the first nine months.

Reference

Locke, E. A., and Latham, G. P. (2002). Building a practically useful theory of goal setting and task motivation: A 35-year odyssey. American Psychologist, 57(9), 705 to 717.