Skip to main content

Prompt: Design a Performance-Critical Lead Funnel

Prompt​

Design a high-traffic paid acquisition lead funnel with campaign variants, experimentation, analytics, trust content, and a multi-step form. Mobile performance directly affects acquisition cost.

Clarifying questions​

  • What are the target Core Web Vitals by cohort?
  • Which content is static, personalized, or experiment-specific?
  • Which scripts are required before conversion?
  • What attribution and consent rules apply?
  • What happens if analytics or experiment scripts fail?

Model answer outline​

  • Static or cached shell for stable content.
  • Edge assignment for campaign and experiment variants.
  • Small hydrated form island with preserved state and idempotent submit.
  • Third-party scripts delayed, owned, budgeted, and kill-switchable.
  • Route budgets for LCP, INP, CLS, JS, images, fonts, and scripts.
  • RUM segmented by campaign, variant, device, network, and release.

Architecture diagram​

Weak answers​

  • Client-side experiment assignment causing flicker.
  • Analytics blocks first render.
  • No third-party owner.
  • No form retry/idempotency.
  • Performance evaluated only by Lighthouse once.

Scoring​

Strong answer: route budget, edge/cache safety, form lifecycle, script governance, RUM segmentation, rollback path, and experiment attribution safety.

For a complete calibrated answer, read worked-answer-performance-lead-funnel and compare it with capstones/capstone-solution-pack-performance-lead-funnel.