Your Sprint Goals Are Embarrassingly Vague. The Research Is 35 Years Old.
Locke and Latham’s goal-setting theory has survived hundreds of replication studies since 1990. The finding is simple and it keeps holding: specific, difficult goals produce better performance than vague “do your best” goals. But only when paired with feedback. Goals without feedback are wishes. Feedback without goals is noise.
Now look at your last three sprint goals.
”Complete Sprint 14 Items”
This is not a goal. It’s a container. It names a bucket of work, not an outcome. If the team finishes 9 of 14 items, did the sprint succeed? Nobody can tell you, because “14 items” was never the real objective. It was a count.
Goals improve performance through four mechanisms. Each one collapses when the goal is vague:
Direction. A good goal points attention at what matters and away from what doesn’t. “Ship saved payment methods for returning users” directs attention. “Complete Sprint 14 items” points at a Jira column.
Effort. Difficult goals increase effort. But only if you can tell whether effort is closing the gap. Vague goals can’t send that signal.
Persistence. Specific goals extend persistence. When you know exactly what done looks like, obstacles are problems to solve. When the goal is fuzzy, obstacles become reasons to stop.
Strategy. Goals trigger strategy. “How do we ship saved payment methods in two weeks with this dependency?” produces useful conversations. “How do we complete 14 items?” produces overtime.
The Feedback Half
The underappreciated half is feedback. Mid-sprint check-ins matter more than retros. The retro is the postmortem. The mid-sprint check lets the team adjust before it’s too late.
At Comcast, I replaced vague sprint containers with explicit sprint goals and biweekly stakeholder demos. Every ten working days, the team knew whether they’d hit the mark. Sprint goal attainment hit 98%. Not because people worked harder. Because ambiguity collapsed.
What About AI Agents?
Agents make goal clarity more important, not less. An agent told to “complete this ticket” will complete the ticket — whether or not it was the right thing to build. An agent told to “reduce checkout abandonment for returning users by enabling saved payment methods” might flag that the ticket misses an edge case.
Vague goals produce vague output. Locke and Latham knew it in 1990. Most sprint backlogs still haven’t caught up.
If your sprint goals start with “Complete” and end with a number, they’re failing the team. Send me your last three. I’ll tell you what’s wrong. s@spencervaradi.com.