Weighted Scoring Model

Choose criteria that represent distinct reasons to invest, such as customer impact, strategic fit, and confidence. Avoid counting the same benefit under several names. Decide how cost will enter the decision rather than adding raw person-days to benefit ratings.

Use a common direction: a higher rating should always mean a better result. If you include delivery ease on a five-point scale, define five as easier delivery and one as harder delivery. Keep the actual effort estimate available separately.

Suppose customer impact has weight 0.5, strategic fit 0.3, and confidence 0.2. A candidate rated 4, 3, and 2 scores (4 × 0.5) + (3 × 0.3) + (2 × 0.2) = 3.3 out of 5. Another rated 3, 4, and 4 scores 3.5.

Those hypothetical results show how the chosen weights affect ordering. The 0.2 difference is not proof that the second option is better when the ratings themselves are uncertain.

Adjust uncertain ratings within a plausible range and see whether the ranking changes. If a one-point change reverses the order, investigate that assumption before making a large commitment.

Use the model to structure discussion, then account for dependencies and mandatory work. Review the weights when strategy changes rather than modifying them quietly to justify a single candidate.