Marketing Copy Board — 3-Lens Ad Copy Review
Review ad, landing page or marketing email copy before it ships: unanswered customer objections, weak CTAs, and risky claims, with cross-checked findings. Reviews ad copy, landing page copy, or marketing email copy using three independent lenses: a Skeptical Customer, a Conversion Realist, and a Brand Guardian. Each lens flags issues blind, then the lenses cross-examine each other, and a board verdict tags every finding CONFIRMED, CORROBORATED, or DISPUTED, ending with a ship recommendation. It quotes the actual copy, avoids inventing product claims, prices, or features, and asks for audience or offer context if none is given. It does not run A/B tests, measure real conversion, or give legal advice; it reviews the text you paste, so redact pricing or unreleased details you would rather not share.
Example input
Use the Marketing Copy Board on this (a sample I wrote). It's the hero section of a landing page for a fictional subscription product, SnoozeBox, a $29/month sleep-tracking ring with an app. Audience: adults 30-50 who sleep badly and have tried apps before. Goal metric: free-trial signups. Headline: The Best Sleep Tracker Ever Made. Revolutionary. Subhead: SnoozeBox uses breakthrough AI to fix your sleep in just 7 days, guaranteed. Our users never wake up tired again, and we beat every other wearable on the market, 100%. Body: SnoozeBox is a lightweight titanium ring with 14 sensors, a 9-day battery, and a companion app with sleep scores, trend charts, and weekly reports. It syncs with all major health apps. Designed in Portland by a passionate team of sleep enthusiasts. Button: Learn More Footer line: $29/month after trial. Cancel anytime via example.com/cancel.
Example output
Context check: audience (adults 30-50, tried apps before), offer ($29/month, free trial) and goal (trial signups) were provided, so the findings below are grounded in those.
## Phase 1 — Independent findings
### 🕵️ Skeptical Customer
1. "The Best Sleep Tracker Ever Made. Revolutionary." No proof point, no reviews, no test, no comparison. Consequence: a stranger who has tried apps before reads this as the same hype they have already seen. Rewrite: "A ring that shows what changed on the nights you slept badly."
2. "fix your sleep in just 7 days, guaranteed" and "never wake up tired again." The first reader thought is "prove it," and nothing answers it. Consequence: the claim destroys trust in the rest of the page. Rewrite: "Most people see their first useful pattern within a week."
3. "breakthrough AI" is undefined. Consequence: sounds like a buzzword; no explanation of what it actually does.
4. The key objection for this audience ("another gadget that ends up in a drawer") and the cost objection ($29/month, which only shows in the footer) are not answered up top. Consequence: a wary visitor leaves before finding out there is a trial and a cancel link.
### 📉 Conversion Realist
1. Button: "Learn More." Consequence: the page's goal is trial signups, yet the only action is vague and low-commitment. Rewrite: "Start my free trial."
2. The headline and subhead describe superiority claims, not the outcome for the reader (what they will see, feel, or do differently). Consequence: skimmers in 8 seconds learn the product claims to be "best," not what it does for them. Rewrite: "Wake up knowing why you slept badly."
3. Concrete specs are buried in the body ("14 sensors, a 9-day battery") instead of leading with benefits. Consequence: features without the "so what."
4. Trial length and price are in the footer only, and there is no reason to act now. Consequence: no urgency or clarity that the trial is free.
### 🛡️ Brand Guardian
1. "fix your sleep in just 7 days, guaranteed." A guarantee of a health result with a timeline. Consequence: a customer screenshots it and demands a refund; also a regulatory-claim risk for a wellness product. Rewrite: remove "guaranteed" and the 7-day promise.
2. "never wake up tired again" and "100%." Absolute language, one counterexample away from a complaint. Consequence: easy to disprove; a single user with insomnia contradicts it.
…Real output from running the skill on a sample problem, shortened. The skill file itself is not published here.