How to Set Project Goals

By Theta PM Systems · Updated November 2026

Most project goals are either restatements of the work ('ship version 2.0') or wishes with no measure ('improve the experience'). Neither helps anyone decide what to do on a Tuesday.

A goal earns its place by changing a decision. If nobody would do anything differently because of it, it is a slogan. The practical version has three parts: an outcome, a measure, and a date — plus a clear line down to the milestones and tasks underneath.

Outcome, Not Output

Outputs are things you ship. Outcomes are things that change because you shipped them. Both matter, but only outcomes tell you whether the project worked.

OutputOutcomeMeasure
Launch onboarding flowNew users reach their first task unaided% completing setup in week one
Rewrite the APIIntegrations stop timing outp95 latency, support tickets per week
Ship reportingLeads review delivery without a manual deckReports opened weekly

The Three Parts of a Usable Goal

A goal that survives contact with a busy quarter has a measure you can check and a date you can miss. Without both, it cannot fail — and a goal that cannot fail cannot guide anything.

  • Outcome: the change in the world, stated plainly.
  • Measure: one number that moves if the outcome happens, with a baseline and target.
  • Date: when you will judge it, not when it will feel finished.
  • Owner: one person accountable for the goal, separate from the people doing the tasks.

From Goal to Milestone to Task

Goals fail when they float above the work. The chain that keeps them connected is short: a goal needs a small number of milestones, and each milestone needs linked tasks that constitute it.

  • One goal per project, at most two. More than that means the priorities are not decided.
  • Three to five milestones, each a decision or a delivery rather than a percentage.
  • Link the tasks that must finish for each milestone, so progress is derived rather than asserted.
  • If a task cannot be traced to a milestone, ask why it is in the project.

Leading and Lagging Measures

Outcome measures are lagging: they tell you whether you succeeded, after the fact. That is too late to steer with. Pair each lagging measure with a leading one you can move this week.

Lagging (outcome)Leading (steering)
Activation rateSetup completion within 24 hours
Support tickets per weekKnown issue count in the backlog
On-time deliveryMilestones with slippage flagged more than a week early

Reviewing Goals Without Turning Them Into Status Meetings

Check goals on a fixed cadence — monthly is usually right — and ask three questions: is the measure moving, is the assumption still true, and what would we stop doing? That last question is the useful one, and it is the one most reviews skip.

If a goal is no longer worth pursuing, say so explicitly and close it. A silently abandoned goal teaches the team that goals are decoration.

Frequently Asked Questions

What makes a good project goal?+

A specific outcome, one measure with a baseline and target, a date you will judge it by, and a single owner. If you could not tell whether it succeeded, it is not yet a goal.

What is the difference between a goal and a milestone?+

A goal is the change you want and how you will measure it. A milestone is a checkpoint on the way there, with a date and linked tasks. Goals are judged; milestones are reached.

How many goals should a project have?+

One, occasionally two. A project with five goals has no priority, and the team will optimise for the one their manager mentions most.

Should goals be SMART?+

The useful parts of SMART are specific, measurable, and time-bound. The rest often produces goals that are achievable and irrelevant, so start from the outcome you actually need.

What if a goal becomes irrelevant mid-project?+

Say so and close it with a note explaining why. Quietly abandoning a goal teaches the team that goals are for show, which is hard to reverse.

Keep Reading

Start free in minutes

Create a workspace, invite your team, and plan your first project today.