Context
Most conversion problems are obvious once you audit systematically — buried CTAs, friction-heavy forms, missing trust signals, slow pages. This skill provides the structured checklist that catches the issues analytics dashboards miss, with fixes prioritized by their expected revenue impact.
Procedure
- Audit CTA visibility: is the primary CTA above the fold? Is it the most visually prominent element? Does the label use action + outcome language? Check on both desktop and mobile.
- Audit form experience: count fields (fewer is better), check for inline validation, error messages, autofill support, mobile keyboard types, and progress indicators for multi-step forms.
- Audit trust signals: look for social proof (logos, testimonials, case studies), security badges, guarantees, and transparent pricing. Score by position and prominence.
- Audit page performance: measure LCP, CLS, INP using PageSpeed Insights. Flag any metric in "poor" range as a conversion blocker.
- Audit mobile experience: check touch target sizes (44px minimum), text readability (16px minimum), form usability on mobile keyboard, and horizontal scroll issues.
- Score each dimension 0-10, calculate the overall conversion readiness score, and produce a prioritized fix list sorted by impact ÷ effort.
Output Format
# Conversion Audit — [URL]
## Scores
| Dimension | Score | Status |
|-----------|-------|--------|
| CTA Effectiveness | /10 | [pass/warn/fail] |
| Form Experience | /10 | [pass/warn/fail] |
| Trust Signals | /10 | [pass/warn/fail] |
| Page Performance | /10 | [pass/warn/fail] |
| Mobile Experience | /10 | [pass/warn/fail] |
| **Overall** | **/50** | **[grade]** |
## CTA Audit
- [ ] Primary CTA visible above fold (desktop): [yes/no]
- [ ] Primary CTA visible above fold (mobile): [yes/no]
- [ ] CTA label uses action verb: [yes/no] — current: "[text]"
- [ ] CTA contrast ratio > 3:1 against background: [yes/no]
- [ ] Secondary CTA offers lower commitment: [yes/no]
- **Finding**: [specific issue]
- **Fix**: [specific recommendation]
- **Impact**: [high/medium/low]
## Form Audit
- [ ] Field count: [number] (benchmark: < 5 for leads)
- [ ] Inline validation present: [yes/no]
- [ ] Error messages are specific: [yes/no]
- [ ] Mobile keyboard types set: [yes/no]
- [ ] Autofill attributes present: [yes/no]
- **Finding**: [specific issue]
- **Fix**: [specific recommendation]
## Trust Signal Audit
- [ ] Customer logos visible: [yes/no]
- [ ] Testimonials with attribution: [yes/no]
- [ ] Case studies / results: [yes/no]
- [ ] Security / privacy indicators: [yes/no]
- [ ] Pricing transparency: [yes/no]
## Performance Audit
| Metric | Value | Threshold | Status |
|--------|-------|-----------|--------|
| LCP | [value] | < 2.5s | [pass/fail] |
| CLS | [value] | < 0.1 | [pass/fail] |
| INP | [value] | < 200ms | [pass/fail] |
## Prioritized Fix List
| # | Fix | Impact | Effort | Priority |
|---|-----|--------|--------|----------|
| 1 | [specific fix] | High | Low | P0 |
| 2 | [specific fix] | High | Medium | P0 |
| 3 | [specific fix] | Medium | Low | P1 |
QA Rubric (scored)
- Specificity (0-5): every finding includes a concrete, implementable recommendation.
- Prioritization (0-5): fixes are ranked by impact-to-effort ratio with clear P0/P1/P2 levels.
- Coverage (0-5): all 5 dimensions audited on both desktop and mobile.
- Evidence grounding (0-5): scores reference benchmarks, data, or best practices — not opinions.
Examples (good/bad)
- Good: "CTA audit: primary button says 'Get Started' but is below the fold on mobile. Fix: move above fold, change to 'Start free scan — no signup required.' Impact: high (mobile is 62% of traffic). Effort: low (CSS + copy change)."
- Bad: "The website could be better. Try improving the user experience and adding more content."
Variants
- Quick scan mode: 10-point checklist focusing on the 3 highest-impact dimensions only — completable in 15 minutes.
- Deep analysis mode: includes heatmap analysis, session recording review, funnel drop-off analysis, and A/B test recommendations for each finding.