skills/audit-premium-digital-experience/references/scoring-and-prioritization.md

# Scoring and Prioritization

## Scoring scale

Use 0–10 only for categories with sufficient evidence:

- **0–2 — dysfunctional:** major failure or unusable in the observed scope.
- **3–4 — weak:** substantial problems obscure value or task completion.
- **5–6 — competent:** functional and credible, but ordinary or inconsistent.
- **7–8 — premium:** polished, coherent, distinctive, and effective with limited gaps.
- **9 — exceptional:** rare integration of concept, craft, usability, content, and technology.
- **10 — reference-level:** near-flawless within the audited scope; use extremely rarely.

Give each score a one-sentence rationale and confidence. Never score unobserved aspects as zero. Mark them `Nicht prüfbar` and exclude them from aggregation.

## Overall score

If enough categories are observable, calculate or reason from these default weights:

| Category | Weight |
|---|---:|
| First impression and positioning | 15% |
| Brand experience | 15% |
| Visual design | 15% |
| UX and conversion | 20% |
| Content | 10% |
| Motion and interaction | 10% |
| Imagery and video | 5% |
| Technical quality | 10% |

Redistribute unavailable weights proportionally among observable categories and disclose that adjustment. Do not manufacture mathematical precision; round to one decimal at most.

## Finding severity

- **Critical:** blocks core use, causes serious accessibility/trust failure, or creates material technical risk.
- **High:** strongly harms comprehension, conversion, brand credibility, or a primary journey.
- **Medium:** meaningful inconsistency or friction with moderate impact.
- **Low:** refinement that improves craft but does not materially change outcomes alone.

## Priority formula

Prioritize using all five dimensions:

- user/business impact;
- evidence confidence;
- implementation effort;
- dependencies;
- regression or brand risk.

Present action bands:

1. **Sofort ändern:** high-confidence, high-impact defects and quick structural wins.
2. **Danach verbessern:** medium effort or dependency-bound improvements.
3. **Premium-Upgrade:** differentiating craft, original interaction, richer narrative, and advanced production after fundamentals are sound.

Use effort labels `S`, `M`, `L`, `XL` only as relative estimates unless the implementation context is known. Do not convert them into money or calendar promises without evidence.