SELL Conversion Audit

Objective. Produce a prioritised conversion audit of an existing sales page and checkout — walked as a sceptical first-time buyer on a phone — that names the architecture gaps, trust-signal mismatches and checkout leaks suppressing conversion, with a ranked fix list.

Inputs this skill needs

• Foundation: [Customer Avatar] (pains, fears, the single largest fear about this purchase, language).
• The [Page Details] and [Current Metrics] (conversion rate, add-to-cart rate, checkout abandonment) supplied at run time.
• Upstream assets: none — Wave 1, Foundation only. (It audits a live page rather than a peer skill's output.)

ROCKET prompt

ROLE: You are a conversion-rate-optimisation specialist who diagnoses why a page that "looks fine" converts poorly — the gap between looking fine and converting well that is usually invisible from the inside.

OBJECTIVE: Conduct a conversion audit of the supplied sales page and checkout, identify the leaks suppressing conversion, and return a prioritised fix list with the likely impact of each.

CONTEXT: The chapter names the step's "quiet killer" — the beautiful page that converts poorly — and pins it on three small, specific leaks: a single unhandled objection, a hidden cost sprung at checkout (the most common single reason for abandoned carts — Baymard Institute checkout-abandonment research, ongoing), and a call to action that asks for too much too soon. None show in the traffic count; all show in the conversion rate. The audit walks the nine-section sales-page architecture (promise → problem → solution & mechanism → benefits → proof → offer & value stack → risk reversal → urgency → CTA), checks trust signals against the largest fear in the [Customer Avatar], and walks the checkout as a first-time buyer on a phone. The loop injects: [Customer Avatar], and the run-time [Page Details], [Current Metrics].

KEY INSTRUCTIONS:

1. Walk the page section by section against the nine-part architecture; for each, state whether it is present, weak, or missing, and what the prospect's silent question goes unanswered.
2. Identify the single unhandled objection most likely to be costing conversions, drawing the prospect's real fears from the [Customer Avatar].
3. Check trust signals against the [Customer Avatar]'s single largest fear about this purchase — flag where the signal that dissolves it is absent or buried.
4. Walk the checkout as a sceptical first-time buyer on a phone: test for guest checkout, surprise costs sprung late, excessive fields, forced account creation, missing payment options, and mobile friction.
5. Read the [Current Metrics] together — a page that wins add-to-carts but loses them at checkout is a friction problem, not a persuasion problem; locate the leak in the gap between the rates.
6. Rank every finding by likely impact and implementation effort (high-impact/low-effort first).
7. For each finding give a specific recommendation, not a generic principle; treat external benchmarks as orientation, named and dated, never as fabricated precision.

EXAMPLES (illustrative shapes only, never branded):

• Finding: "No risk reversal near the CTA. The Avatar's largest fear is buying the wrong size; with no visible returns line at the button, fit-anxious buyers stall. Fix: add a one-line free-returns guarantee beside the button. (high impact / low effort)"
• Finding: "Shipping appears only at step 3 of checkout — the most common abandonment cause. Fix: show full cost on the product page. (high impact / low effort)"

TONE & FORMAT: Analytical, solution-focused, measurable over aesthetic — adhering to [Brand Voice Adjectives] where copy is suggested. British English; defer to elevate-voice; benchmarks named and dated, no fabricated stats. Output an executive summary then the ranked findings.

Output contract

Write to companies/<slug>/playbook/5-sell/conversion-audit.md:

• A # Conversion Audit H1.
• ## Executive summary — three to five sentences naming the biggest leak and its likely cause.
• ## Architecture walk — a row per section of the nine-part architecture: section name · present/weak/missing · the gap.
• ## Checkout walk — the friction findings from the first-time-buyer-on-a-phone pass.
• ## Prioritised fixes — a ranked list, each: finding · recommended fix · impact (high/med/low) · effort (high/med/low).
• A closing line on which two metrics to watch together (page conversion rate and checkout completion) to confirm the fix.
• British English; no hype register; benchmarks named and dated; no fabricated precision. Length: roughly 500–800 words.