Page Conversion

---
name: page-conversion
description: Audit and repair conversion on marketing pages, lead and contact forms, and the overlays attached to them: message match, offer clarity, proof placement, form fields and errors, popups, sticky bars, slide-ins, and exit interventions, with instrumentation and an experiment-ready hypothesis. Use for weak page conversion, form abandonment, or URL-based conversion reviews; use activation-path for account creation and onboarding, and upgrade-moments for in-product paywalls.
license: MIT
metadata:
  version: "2.0.0"
  author: MTEnt
---

# Page Conversion

Diagnose the weakest evidenced transition on the page. A conversion rate shows an outcome, not its cause, and an interruption must earn attention through relevance.

If `.agents/marketing-context.md` exists, read it first; it is the truth file every skill in this hub shares, and its evidence states are constraints. Read [page-and-form-workflow.md](references/page-and-form-workflow.md) for the audit, page types, form behavior, experiment design, and output. Read [overlay-workflow.md](references/overlay-workflow.md) when the surface includes a popup, banner, modal, or exit intervention.

## Operating contract

1. Define page job, audience, source, offer, eligible population, action, period, baseline, device mix, and downstream quality.
2. Inspect the rendered experience and working path when accessible; separate observations from causal hypotheses and competing explanations.
3. Repair broken behavior and accessibility before persuasion experiments.
4. For overlays choose the least disruptive format that can do the job and specify trigger, dismissal, suppression, frequency, conversion exit, and mobile behavior.
5. Produce complete copy, layout or component specifications, or an experiment-ready hypothesis as requested, with events and guardrails.

Do not use dark patterns, fake proof, hidden costs, disguised close controls, preselected consent, or friction removal that destroys lead or customer quality.

## Return

Inspected surface and evidence, observed friction separated from hypothesis, prioritized changes, requested copy or interface artifact, event plan, experiment or monitoring plan, and unverified behavior.