Strategy should remove work
A North Star is not useful because it sounds directional. It becomes useful when it changes what receives attention, funding, and leadership time. If every initiative can claim alignment, the strategy has not created a choice. The first test is whether teams can explain which work should stop, wait, combine, or move faster.
Create a common intake
Projects often enter through different doors with inconsistent evidence. A shared intake should clarify the problem, beneficiary, expected outcome, strategic driver, owner, dependencies, cost, risk, urgency, and proof of value. The aim is not bureaucracy. It is comparability before influence and momentum make the decision.
Separate value from confidence
An initiative can promise high value with weak evidence. Score expected value and confidence separately so uncertainty remains visible. This creates room for discovery, pilots, or staged investment instead of forcing a premature yes or no. It also prevents precise-looking scores from hiding assumptions.
Use criteria that reflect the strategy
Generic prioritization methods are a starting point, not the final model. Weight criteria according to the North Star and current constraint: customer impact, growth, cost, risk reduction, capability, urgency, reach, feasibility, data readiness, or learning value. The model should reveal strategic trade-offs rather than average them away.
Prioritize the portfolio, not isolated ideas
The strongest individual projects can still create a weak portfolio if they compete for the same people, depend on unfinished foundations, or over-concentrate risk. Review initiatives together. Look for sequencing, capability bottlenecks, duplicated outcomes, dependencies, and the balance between foundational work, near-term value, and strategic options.
Make decisions and assumptions visible
Record why an initiative moved, paused, or stopped; which evidence mattered; and what would change the decision. Transparency helps teams improve future submissions and prevents the same debate from restarting without new information. It also turns prioritization into organizational learning rather than a periodic ranking event.
Revisit when evidence changes
Priority is not a permanent identity. Outcomes, constraints, costs, dependencies, and strategic conditions move. A recurring portfolio review should compare expected and actual evidence, refresh capacity, and deliberately reallocate attention. Operational excellence is partly the discipline to stop protecting yesterday's priority.