A target can make a difficult project easier to coordinate. It can also keep you working on the wrong problem after the facts change. Both things can be true of the same target at different times. This guide looks at how to choose the right kind of goal, how to see what the target is doing to your decisions, and when to revise or leave it.
Decide what kind of decision the goal should help you make

A broad direction helps you choose between opportunities. “I want to do more creative work with other people” can guide a decision about taking a new role, joining a project, or freeing up a weekend. It leaves room for different ways to pursue the same value. A specific target helps you coordinate an attempt. “By the end of May, run two feedback sessions on the prototype” tells a team what to do and when to review the result. You may need both, especially when a project is only one way to express a longer-term priority.
Edwin Locke and Gary Latham’s review of goal-setting research found that specific, challenging goals often improved task performance compared with vague instructions to do one’s best. The review also discusses factors that shape the effect, including commitment, feedback, ability, and task complexity. It is useful evidence for choosing a performance target when the task is understood. It leaves a separate question about whether the target serves your priorities. [1]
Use a performance goal for a task with a known method
Consider a designer who can already prepare a usable prototype and organise interviews. A target of running five sessions by a review date helps the designer schedule work and coordinate with colleagues. Specify what evidence each session should produce. Five calls where participants only watch a demonstration would answer a different question from five attempts to use the prototype unaided. Ask the team to agree on the question before setting the count.
A demanding target can focus attention when there is a credible route to the result. Count the time and access the work needs. If the designer can contact only two eligible participants this month, the original target depends on solving recruitment. The next milestone might be to confirm access to a suitable group, and the interview date can follow that decision.
Use a learning goal while the strategy is still uncertain
An inexperienced interviewer may get more value from trying two ways to ask open questions and checking which yields clearer examples. A target based on the number of interviews alone could encourage them to repeat the same poor method. Seijts and Latham distinguish tasks that need effort and persistence from tasks that first need effective strategies or skills. A specific learning goal can direct effort toward finding a strategy. [2]
State what you want to learn and what comparison would help. In the example, choose two interview openings, record where people need clarification, and compare whether they give concrete examples. When the interviewer can repeat a technique that works, the next goal can set a performance standard. A useful sequence moves from evidence about the method to evidence about the result.
Connect the goal to a reason you endorse

Imagine two people pursuing the same promotion. One wants the chance to mentor others and sees a new role as a way to do that. The other likes their current work but feels behind their peers. The title alone cannot tell you which path fits either person. In longitudinal research on self-concordance, Sheldon and Elliot found that goals more consistent with people’s interests and values attracted more sustained effort and were more likely to be attained. A reason can be personally endorsed even when a goal begins with an external requirement. [3]
Try writing two lines under a goal. “I want this because ___.” “If this route becomes unavailable, I could still serve that reason by ___.” Someone who wants to mentor might start a peer review group or coach a new colleague even if promotion is delayed. Someone who wants a specific salary change may identify different routes. Naming the reason gives you options when the route becomes expensive or unavailable.
Check the opportunity cost before you commit
Look at the next two weeks and ask which work will have to move. Adding a demanding goal on top of existing commitments can turn a sound idea into an impractical promise. Ask what you are willing to defer and what you need to protect. A learning project might be worth two evenings and one weekend morning. It may stop being a good choice if it requires every evening for two months. A resource boundary lets you check the trade-off before circumstances make it for you.
Inspect what your metric rewards
A target tells people where to look. It also changes where they spend time. A service team asked to close tickets quickly may send short replies that increase repeat contacts. A writer trying to meet a weekly word count may finish pages that still leave the argument unclear. Ordóñez and colleagues’ critique of overly prescriptive goals describes risks such as narrowed attention and neglected dimensions of performance. Treat those risks as questions to ask about your own metric. [4]
Write a quality constraint beside the target. For support, pair time to first response with a check on reopens or resolution quality. For writing, ask a reader to state the argument before evaluating the word count. Include an example of what would count as a false success. This is particularly important when a target will affect compensation, grades, or other people’s workload. Revisit the measure if it starts to reward behaviour that undermines the original purpose.
Keep an early signal and an outcome
An action measure shows whether the plan happened. An outcome measure tells you whether the situation improved. Interview invitations sent, participants recruited, and usable findings answer three different questions. If the invitations went out and nobody agreed to join, inspect the recruitment message or access to the intended group. If interviews happened but yielded little evidence, review the questions or tasks used during the sessions. Use each measure to locate the next decision.
A meta-analysis led by Harkin found that increasing progress monitoring helped goal attainment on average. A measure earns its place in a plan when it helps you make a better choice. If the team checks three numbers every week but only one ever changes the next action, simplify the review. [5]
Choose a measure that can tolerate normal variation
A weekly number can move because of case mix, holidays, or a small denominator. Suppose a support team resolves 9 of 10 cases in one week and 40 of 50 the next. The percentages look far apart, but each number comes from a different amount of work and may cover different kinds of requests. Show the counts, compare similar cases, and look for a pattern before replacing a strategy after one noisy week. If a particular complaint describes serious harm, investigate that case directly.
Qualitative goals have a related problem. Two reviewers may mean different things by “clear.” Agree on a small set of observable criteria, then compare one example together. If a reader can find a recommendation but cannot explain what it depends on, the review points to a specific weakness. You can keep that criterion even if the document format changes.
A useful test for any metric is to ask what you would do if it improved while the problem remained. If the answer is “keep celebrating the number,” give the underlying problem its own check before the target starts driving decisions.
Decide how you will respond to new evidence
Set a review date and write a few conditions in advance. You might keep going if the next two attempts produce comparable observations, change the method if both fail for the same fixable reason, or pause if a necessary approval disappears. Conditions written early give you something to consult once you have invested time and are tempted to keep the original plan by default.
- Continue when the aim is still useful, the route remains feasible, and the evidence is moving in the intended direction.
- Revise when the aim still matters but the method, timeline, or measure needs adjustment. Change the part responsible for the gap.
- Pause when an essential resource is missing and you can say what would justify restarting.
- Close when the target is no longer feasible or worthwhile. Record what you learned and consider whether the underlying reason has another route.
Wrosch and colleagues found associations between being able to disengage from unattainable goals, reengage with alternatives, and subjective well-being. Their findings support taking flexibility seriously when a goal has become unattainable. A difficult week alone gives you little information about that condition. Look at the constraints, the quality of your attempts, and the alternatives available. [6]
A worked decision after a missed target
Suppose the designer intended to run five sessions, managed two, and found the same serious navigation problem in both. The missed count leaves uncertainty about how widespread the problem is. The observations still give the team a reason to fix an obvious obstacle before asking three more people to use the same version. Record the limitation, make a targeted change, and use the remaining sessions to test the revised flow. Keep the goal of improving the prototype. Update the plan to reflect what the work has shown.
Now suppose the product is cancelled. The interview target has lost its purpose. The designer can save the interview notes, record which prompts worked, and apply the technique to a project with an actual decision ahead. That is a different response to the same missed count. The right choice depends on what changed and what the goal was meant to serve.
Questions for the next goal review
What decision is this goal helping me make? What would look like progress? Which assumption has changed? What does the target leave out? Given the cost and what I know now, would I choose the same next action?
For a step-by-step routine to put those decisions into practice, see the eight-step goal-setting cycle.
References
- Locke & Latham (2002). Building a practically useful theory of goal setting and task motivation.
- Seijts & Latham (2005). Learning versus performance goals.
- Sheldon & Elliot (1999). Goal striving, need satisfaction, and longitudinal well-being.
- Ordóñez et al. (2009). Goals gone wild. Harvard Business School working paper 09-083.
- Harkin et al. (2016). Does monitoring goal progress promote goal attainment?
- Wrosch et al. (2003). Adaptive self-regulation of unattainable goals.
Further reading
Read the original guide from Mark Manson.

