Setting goals that measure the outcome you want and actually steer what you do.
A goal is only useful if it can change a decision. That sounds obvious and rules out most of what gets written down at work: goals that measure activity instead of results, goals nobody looks at between January and December, goals that never become anything you'd do on a Tuesday. A good one names the outcome you want, has a way of telling whether you got there that you can't quietly reinterpret, and gets reviewed often enough to steer. This is a longer horizon than day-to-day prioritization — that's a separate problem of ordering and slack — but the two connect at exactly one point: a goal that never wins against the urgent work isn't a goal, it's a preference.
Sets goals that measure effort, because effort is the part that can be controlled. Attend twelve workshops. Publish twenty posts. Hold thirty customer conversations. Every one is met, in full, on time, and the year ends with the underlying situation exactly as it was, since nothing in the goal ever required it to change. Reports complete attainment against all objectives, which is true, and files it as a good year on the available evidence.
Writes a thoughtful set of goals each January, in a document of real quality, and then does not open it. Rediscovers the document in early December while preparing for the review, reads it with the mild interest of an archaeologist, and finds that four were achieved by accident and three describe a strategy the company abandoned in March. Writes a new set for the following January, thoughtfully, in a document of real quality, and does not open it.
Genuinely wants all of it — the certification, the side project, the move into management — and has never converted any of it into a next action or a slot in an actual week. The wanting is real and recurring, surfacing most Sunday evenings. Has held the same three aspirations, unchanged and unstarted, for long enough that they now function less as goals than as a stable feature of their personality.
Adjusts the definition as the year proceeds, always slightly, always with a defensible reason, so that the goal and the outcome converge by December. Ends every cycle having met every target, which produces an unbroken record and destroys the only thing goals are good for, which is finding out where your estimate of yourself was wrong. Nothing is ever learned, because nothing is ever missed, and the record grows more impressive and less informative each year.
Your goal for the quarter is "publish 20 blog posts" and you're on track. What would you do — keep going, or check whether the twenty posts are producing whatever you wanted them for, and change the measure if not?
It's October and you haven't looked at your goals since they were set. What would you do — quietly rewrite them to match what you actually did, or read them as written and note honestly which ones you missed and why?
You want to move into management "at some point." What would you do — keep it as an intention and watch for opportunities, or name the two concrete things you'd need to have done by June and put them in a week?