Generated from the gaps you reported. Severity and weak-P weighted.
You answer 18 short diagnostic questions across the seven Ps. Each answer carries a weight (strong, okay, weak, missing). Your mix score is the weighted percentage of points earned, so it is a genuine 0 to 100. Per-P scores tell you which P is dragging the average. Recommendations are produced only for the questions you answered weak or missing, and ordered by severity plus how much the underlying P is below the average. Nothing is invented: every recommendation maps to a question you answered honestly.