Landing Page Friction Audit — Performance Marketing Automation Workflow

schema_version: 1
version: "1.0.0"
id: friction_audit_engine
title: "Landing Page Friction Audit"
description: "Score a landing page across 7 friction types using severity-weighted rubric. Outputs prioritized fix list with expected conversion lift."
category:
  level: task
  domain: marketing
  reusable: true
params:
  page_description: string
  ad_copy: string
  current_metrics: string
steps:
  - type: llm.generate
    prompt: |
      You are a CRO specialist conducting a friction audit on a landing page.

      **Page description:** {{page_description}}
      **Ad copy driving traffic:** {{ad_copy}}
      **Current metrics:** {{current_metrics}}

      Score the page across 7 friction types. For each, rate severity 1-10 and estimate conversion impact.

      **Friction Types:**
      1. **Navigation friction** — Full site nav, too many links, competing CTAs
      2. **Copy friction** — Vague headline, jargon, no clear benefit, message mismatch with ad
      3. **Visual friction** — Cluttered layout, unclear hierarchy, slow images
      4. **Form friction** — Too many fields, unclear labels, no progress indicator
      5. **Trust friction** — No social proof, no security badges, no privacy assurance
      6. **CTA friction** — Vague button text, below fold, multiple competing CTAs
      7. **Speed friction** — Page loads >3 seconds, render-blocking scripts

      **Message Match Check:**
      Score ad-to-page message match (Strong/Moderate/Weak):
      - Does page headline echo ad headline benefit?
      - Does page offer match ad offer exactly?
      - Is there visual continuity?

      **Output format:**

      ## Message Match Score: [Strong/Moderate/Weak]
      [Specific mismatches identified]

      ## Friction Scores
      | Friction Type | Severity (1-10) | Evidence | Est. CVR Impact |
      [7 rows]

      ## Overall Friction Score: [sum/70]
      [Interpretation: <20 = low friction, 20-35 = moderate, 35-50 = high, >50 = critical]

      ## Prioritized Fixes (by ICE score)
      | Fix | Impact | Confidence | Ease | ICE Score | Est. Lift |
      [Top 5-8 fixes ranked by ICE]

      ## Quick Wins (implement today)
      [Fixes with Ease >= 8]

      ## Test Recommendations
      [Top 3 A/B tests to run, with hypothesis for each]
    as: audit_result

  - type: variable.set
    name: result
    value: "{{audit_result}}"