SMART Goals

Most goals do not fail because the person was lazy. They fail because the goal was never something you could succeed or fail at in the first place. "Get fitter", "be better at my job", "read more" — none of these can be finished, which means none of them can be started either. There is no first action, no way to tell a good week from a bad one, and no moment where you are allowed to say it is done.

SMART is a fix for that specific problem. It is not a productivity system, a planning method, or a theory of motivation. It is a checklist for phrasing — five questions you ask an intention until it turns into something testable. That is a smaller claim than most articles about it make, and it is the reason the framework has survived forty years while heavier systems came and went.

What SMART actually stands for

SMART is an acronym for Specific, Measurable, Achievable, Relevant, and Time-bound. You will see it attributed to George T. Doran, and that is right — he introduced it in a 1981 issue of Management Review, in a paper arguing that managers set objectives too vaguely to be managed.

It is worth knowing that Doran's original letters were not quite the ones in circulation today. He wrote Assignable where we now say Achievable, and Realistic where we now say Relevant. The modern version drifted through decades of corporate training decks, and the drift was mostly an improvement: "assignable" only makes sense for a manager handing out work, while "achievable" and "relevant" are questions anyone can ask about their own goal.

The point of knowing this is not trivia. It is that SMART is a convention, not a law. Nobody validated these five criteria against outcomes. They are five sensible questions that happen to spell a word, and treating them as five sensible questions — rather than a compliance form — is what makes them useful.

The five criteria, one at a time

Each letter is a question, and each has a characteristic way of being answered badly. The failure mode matters more than the definition.

Specific — what exactly will you do?

A specific goal names the action, not the aspiration. The test is whether someone else could read it and know what to do tomorrow morning.

"Get fitter" fails. "Run three times a week, 5km each" passes: the action is named, the volume is named, and there is nothing left to interpret.

The usual mistake is stopping at the domain — "focus on health", "improve communication". A domain is a category of goals, not a goal.

Measurable — how will you know it is done?

Measurable means there is an observation that settles the question. Not a feeling that you are making progress; an observation.

Ask it as: what would I look at? For the running goal, you would look at the log: 15km this week or not. For a writing goal, the word count. For a revenue goal, the number in the dashboard.

The characteristic failure is a measure that sounds numeric but is not observable — "improve customer satisfaction by 20%" is measurable only if you already run a survey that produces a comparable number. If you do not, you have written a goal that needs a second project before it can be checked.

Achievable — what makes this realistic for you, right now?

Achievable is the most misread letter. It does not mean easy, and it does not mean comfortable. It means there is a plausible route from where you are to where the goal is, using resources you actually have.

The honest version of this question is: what would have to be true? Running 15km a week is achievable if you already run 5km a week. It is not achievable in the first week if you have not run in two years — not because the distance is impossible, but because the route from here to there is missing a few months.

Achievable is also the letter that invites sandbagging. A goal you are certain to hit tells you nothing you did not already know, and teams under target-based review learn very quickly to write those. If you are setting a goal that someone else will assess you against, be aware you are also setting the incentive to aim low.

Relevant — why does this matter, and why now?

Relevant asks whether the goal connects to something you actually care about. It is the only one of the five criteria that can invalidate the entire goal rather than just its phrasing.

The test is a chain: if I achieved this, what would be different, and would I want that? Running 15km a week matters because you want to finish a 10k in March without walking. If you cannot finish that chain in a sentence or two, you have probably inherited the goal rather than chosen it.

Time-bound — by when?

A deadline turns a goal into a thing with a shape. Without one, every day is an equally good day not to start, which is exactly what happens.

The failure mode is a deadline so distant that it exerts no force. "By the end of the year" is technically time-bound and practically identical to having no date at all. More on this below — the fix is not a nearer deadline but a rhythm of checkpoints.

Three vague goals, rewritten

The criteria are easier to trust once you have watched them do the work. Here are three intentions of the kind people actually write, each put through the five questions.

"I want to get fitter." Specific: run three times a week, 5km each. Measurable: 15km logged in the running app by Sunday night. Achievable: I already run once a week, so this is two more runs, not a new habit from zero. Relevant: I want to finish the March 10k without walking. Time-bound: every week until 1 March.

Read as a sentence: Run 5km three times a week, every week until 1 March. I will know it is done when the app shows 15km for the week.

"I need to be better at my job." This one is mostly a Relevant problem — "better" is not a direction. Pushed on what would actually change, it becomes: I want to stop being the bottleneck on code review. Specific: review every PR assigned to me within one working day. Measurable: no PR in my queue older than 24 hours at the end of the day. Achievable: I get three or four a week, which is about 40 minutes daily. Relevant: two releases slipped last quarter waiting on me. Time-bound: starting Monday, held for the whole quarter, checked at each retro.

"I should read more." Specific: read 20 pages before opening my laptop. Measurable: pages, marked in the book. Achievable: 20 pages is roughly 25 minutes, which is the length of a single Pomodoro session and already exists in my morning. Relevant: I have bought nine books this year and finished one. Time-bound: every weekday in October, then reassess.

Notice how differently the three behave. The first was a phrasing problem — the intention was fine, the words were vague. The second was a goal selection problem, and SMART only helped after the real objective surfaced. The third turned on Achievable: it worked because it attached the goal to a slot in the day that already existed.

Build one

Work through your own. The five fields assemble into a sentence as you fill them in — if the sentence reads awkwardly, that is usually a real problem with the goal rather than the wording.

Where SMART breaks down

Any framework worth using is worth knowing the limits of, and SMART has three that matter.

It grades phrasing, not worth. A specific, measurable, time-bound version of the wrong objective is still the wrong objective — and now it has the appearance of rigour, which makes it harder to abandon. SMART cannot tell you whether the goal deserves your quarter. Checking that is a different job: first principles thinking is the tool for asking whether the objective is real or inherited.

Measurable pulls toward the trivial. The things easiest to count are rarely the things that matter most, so a strict measurability requirement quietly reshapes goals toward whatever is convenient to observe. Pages read is a worse target than understanding gained; it is just the one with a number. Use the number, but stay aware it is a proxy.

It says nothing about trade-offs. Five SMART goals that each pass the checklist can still be collectively impossible, because the checklist looks at one goal at a time and never asks what you are giving up. If everything on your list is a priority, nothing is — which is the problem the Eisenhower Box exists to solve, and worth running first if you have more candidate goals than weeks.

A useful stress test: ask what would have to go wrong for this goal to fail, and write that down next to it. That is inversion, and it catches the obstacles the five positive questions never think to ask about.

Pick a measure you can act on

The single highest-leverage improvement to most SMART goals is choosing a better thing to measure. There are two kinds, and the distinction decides whether the goal is useful day to day.

A lagging measure reports the outcome you want: revenue, weight, the finished manuscript. It is the thing you actually care about, and it is almost useless as a target, because by the time it moves the behavior that moved it is weeks in the past.

A leading measure reports the behavior that produces the outcome: discovery calls made, sessions run, pages drafted. It is within your control today, and it changes immediately when you change what you do.

State the lagging measure as the reason and the leading measure as the goal. "Draft 500 words every weekday, aiming at a finished first draft by June" gives you something to succeed at daily and something to aim at overall. "Finish the book by June" gives you one verdict, in June.

Give it a review rhythm

A deadline at the far end of a long project is not a constraint, it is a rumour. Nothing about a date in March changes what you do in November. What changes November is a checkpoint in November.

The checkpoints do not need to be elaborate. A weekly look at the leading measure is enough, and it only has to answer three questions.

Reviewing also gives you permission to revise. Goals set in January were set by someone with less information than you have now, and holding one that has been overtaken by events is not discipline — it is just a sunk cost with a deadline attached. Change the goal deliberately at a checkpoint, and write down why. That is a different act from quietly letting it lapse.

If you want the habit to stick, attach it to something that already happens: the same slot each Friday, or the end of a weekly review you already run.

When to reach for something else

SMART is the right tool when you know roughly what you want and the problem is that you have said it too loosely. It is the wrong tool in three common situations.

If you have too many goals rather than vague ones, the constraint is selection, not phrasing — start with the Eisenhower Box or The ONE Thing.

If the goal is clear but the day keeps filling with other things, the problem is defending time, and timeboxing is the answer.

And if you keep hitting your goals and the results still are not what you hoped, the objective itself is probably wrong. Second-order thinking is the move there: not "did I achieve it", but "what happened next, and was that what I wanted?"

Used for what it is — a phrasing checklist, run in about five minutes, revisited weekly — SMART is one of the highest-return things you can do to a goal. Used as a system in its own right, it produces very well-specified goals that nobody achieves.