Prioritization gets easier when trade-offs are visible

Prioritization is a choice between competing uses of attention

Frameworks can help structure a discussion, but they can also create false precision. A feature with a score of 82 is not objectively better than one with 76 if the underlying assumptions are uncertain or the scoring system hides strategic trade-offs.

The useful part of prioritization is making those trade-offs visible.

Start with the outcome, not the item

For each candidate, write the problem, target customer, expected outcome, and why the work matters now. If the item cannot be connected to a meaningful outcome, it is difficult to compare with other work.

This also exposes duplicate ideas that are different solutions to the same problem.

Separate evidence from confidence

List what you actually know: observed behavior, repeated interviews, revenue impact, support volume, experiment results, technical constraints, or contractual commitments.

Then list what is assumed. Confidence should reflect the quality and relevance of the evidence, not the enthusiasm of the person proposing the work.

Make cost broader than engineering effort

Implementation time matters, but so do design, data migration, support load, sales enablement, operational complexity, maintenance, and the attention needed after launch.

A “small” feature that creates permanent exceptions can be more expensive than a larger project with a clean operating model.

Add risk and reversibility

Some decisions are easy to reverse. Others create customer expectations, data structures, pricing commitments, or technical dependencies that are difficult to unwind.

When evidence is weak, reversible experiments deserve a different treatment from irreversible commitments.

Name what will not happen

Every priority displaces something. Write the opportunity cost explicitly: “If we do this in September, onboarding research moves to October.”

This changes the discussion from “Is this a good idea?” to “Is this a better use of our next unit of capacity than the alternative?”

Use scores as prompts, not verdicts

A scoring framework can flag disagreement. If one person gives reach a 9 and another gives it a 3, the useful result is not the average. The useful result is the question: what evidence are we interpreting differently?

Record the decision and revisit condition

After choosing, document why. Include the evidence, assumptions, trade-offs, and what new information would justify revisiting the decision.

Prioritization gets easier when the team can see the real choice, not when the spreadsheet produces a more impressive decimal.