Skip to Content
DocsGoalsOverview
DRAFT — This document is under development and not yet reviewed.

Goals and Objective Setting

Updated 27 March 2026

A scaled delivery organization sets goals at multiple levels. Each level has a different timescale, a different purpose, and a different relationship to value. The levels are not a cascade — they are a system. Understanding how they connect is what makes goal-setting useful rather than bureaucratic.

Three Levels of Goals

Outcome goals operate at portfolio level. They describe changes in the world the organization wants to bring about — and they are the verification layer of the system: the place where the organization asks whether its delivery has mattered. OKR is the dominant model. See Outcome Goals.

PI goals operate at DVS level. They describe working outputs that a DVS commits to having in place by the end of a Program Increment — integrated, runnable, and demonstrable. A PI goal is a value hypothesis: we believe this working output, once used, will contribute to our outcome goals. See PI Goals.

Sprint goals operate at team level. They describe the single working output a team commits to achieving in a sprint. Like PI goals, a sprint goal carries a value hypothesis — verified at Sprint Review through running software. See Sprint Goal.

Alignment, Not Cascade

Goals at lower levels align toward goals at higher levels — they do not mechanically decompose from them.

A portfolio-level OKR does not break down into a set of PI goals that sum to the OKR. DVS teams and product managers form their own judgment about which working outputs, built this PI, are most likely to move the outcomes the portfolio cares about. That judgment is the product of understanding context, constraints, and current knowledge — it cannot be replaced by arithmetic.

The same applies between PI goals and sprint goals. Teams select sprint goals that serve their PI commitments, but the connection is directional, not mechanical.

When goals are treated as a cascade rather than an alignment, the result is false precision: lower levels report progress on decomposed sub-goals that may have no real connection to whether the higher-level outcome is moving.

  • Outcome Goals — OKR model, portfolio level, the verification layer
  • PI Goals — working outputs with value hypotheses at DVS level
  • Sprint Goal — team-level working output commitment for a single sprint
PI PlanningPI Goals