Menu
Audit existing sales funnels against RMBC and CRO best practices — score each step, identify conversion leaks, and output a prioritized fix matrix.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
Based on SOC occupation classification
Upgrade RMBC Skills to the latest version and show what's new.
Generate structured A/B test plans for DTC funnels — hypothesis, control vs variant, primary metric, sample size estimate, test duration, and success criteria using RMBC principles.
Generate 5-7 unique ad angles for paid media campaigns — each a full strategic positioning direction with headline, body, visual, and platform fit.
Audit paid media ad creative (copy + visual direction) against RMBC and DR principles — score hook, mechanism, proof, CTA, and visual-copy alignment with actionable fixes.
Generate RMBC-structured advertorial and native ad copy that reads like editorial content but drives direct response conversions.
Design bonus stacks that increase perceived value without increasing fulfillment cost — 3-5 bonuses with name, description, perceived value, actual cost, and core offer complement rationale using RMBC principles.
| name | funnel-audit |
| description | Audit existing sales funnels against RMBC and CRO best practices — score each step, identify conversion leaks, and output a prioritized fix matrix. |
| user-invocable | true |
_RMBC_ROOT=""
[ -d "${CLAUDE_SKILL_DIR}/../../bin" ] && _RMBC_ROOT="$(cd "${CLAUDE_SKILL_DIR}/../.." && pwd)"
[ -z "$_RMBC_ROOT" ] && for _D in "$HOME/.claude/skills/dtc-copywriting-skills" ".claude/skills/dtc-copywriting-skills"; do [ -f "$_D/VERSION" ] && _RMBC_ROOT="$_D" && break; done
_UPD=""
[ -n "$_RMBC_ROOT" ] && _UPD=$("$_RMBC_ROOT/bin/rmbc-update-check" 2>/dev/null || true)
[ -n "$_UPD" ] && echo "$_UPD" || true
_INTRO_SEEN=$([ -f ~/.rmbc-skills/.intro-seen ] && echo "yes" || echo "no")
_TEL_PROMPTED=$([ -f ~/.rmbc-skills/.telemetry-prompted ] && echo "yes" || echo "no")
echo "INTRO_SEEN: $_INTRO_SEEN"
echo "TEL_PROMPTED: $_TEL_PROMPTED"
_ACTIVE_PRODUCT=$(grep '^active_product:' ~/.rmbc-skills/config.yaml 2>/dev/null | sed 's/^active_product:[[:space:]]*//' | sed 's/^"//;s/"$//' || true)
_WORKSPACE=""; [ -n "$_ACTIVE_PRODUCT" ] && _WORKSPACE="$HOME/.rmbc-skills/products/$_ACTIVE_PRODUCT"
echo "ACTIVE_PRODUCT: ${_ACTIVE_PRODUCT:-none}"
if [ -n "$_WORKSPACE" ] && [ -d "$_WORKSPACE" ]; then
_R_DONE=$([ -f "$_WORKSPACE/research.md" ] && echo "yes" || echo "no")
_M_DONE=$([ -f "$_WORKSPACE/mechanism.md" ] && echo "yes" || echo "no")
_B_DONE=$([ -f "$_WORKSPACE/brief.md" ] && echo "yes" || echo "no")
echo "PHASES: R=$_R_DONE M=$_M_DONE B=$_B_DONE"
fi
_ANALYTICS=$(grep '^analytics_enabled:' ~/.rmbc-skills/config.yaml 2>/dev/null | sed 's/^analytics_enabled:[[:space:]]*//' || echo "true")
[ "$_ANALYTICS" = "true" ] && [ -n "$_RMBC_ROOT" ] && timeout 2 "$_RMBC_ROOT/bin/rmbc-analytics" log --skill "funnel-audit" --product "${_ACTIVE_PRODUCT:-none}" --tier 3 2>/dev/null &
_SESSION_COUNT=$(ls /tmp/rmbc-session-* 2>/dev/null | wc -l | tr -d ' '); touch "/tmp/rmbc-session-$$"
echo "SESSIONS: $_SESSION_COUNT"
If output shows UPGRADE_AVAILABLE <old> <new>: read skills/rmbc-upgrade/SKILL.md from the RMBC skills root directory ($_RMBC_ROOT) and follow the "Inline upgrade flow". If JUST_UPGRADED <old> <new>: read $_RMBC_ROOT/CHANGELOG.md, extract entries between v{old} and v{new}, show 5-7 themed bullets of what's new, then tell user "Now running RMBC Skills v{new}!" and continue.
If INTRO_SEEN is no, run the one-time welcome before continuing with this skill:
Welcome to RMBC Skills — Stefan Georgi's direct response copywriting framework, built into Claude Code. 44 skills covering hooks, ads, emails, landing pages, VSL scripts, and more.
Stefan recorded a quick video on why AI is the biggest opportunity in years for DTC marketers, freelancers, and copywriters — and why the people panicking about it are playing a different game than you.
Use AskUserQuestion:
If "Yes, open the video":
open "https://www.youtube.com/watch?v=zI8tNfefH1M"
mkdir -p ~/.rmbc-skills
touch ~/.rmbc-skills/.intro-seen
If "Skip — let's go":
mkdir -p ~/.rmbc-skills
touch ~/.rmbc-skills/.intro-seen
Continue with this skill immediately.
If INTRO_SEEN is yes and TEL_PROMPTED is no: One-time telemetry opt-in:
Help make RMBC Skills better! Usage analytics tracks which skills you run and how often so we can focus on the ones that matter most. Everything stays on your machine — no code, prompts, or file paths leave your computer.
Use AskUserQuestion:
If "Yes, that's fine":
mkdir -p ~/.rmbc-skills
touch ~/.rmbc-skills/.telemetry-prompted
If "No, turn it off":
mkdir -p ~/.rmbc-skills
touch ~/.rmbc-skills/.telemetry-prompted
sed -i '' 's/^analytics_enabled:.*/analytics_enabled: false/' ~/.rmbc-skills/config.yaml 2>/dev/null || true
Continue with this skill.
Audit an existing DTC sales funnel against RMBC methodology and CRO best practices. Every funnel leaks revenue somewhere — the question is where and how much. This skill evaluates each step in the funnel (traffic → landing → order → upsell → thank you → email) with per-step scoring across five dimensions. Output: a funnel health score (0-100), identification of the top 3 conversion leaks, and a prioritized fix matrix ranked by revenue impact.
| Input | Required | Description |
|---|---|---|
funnel_description | Yes | Detailed description of each funnel step — pages, copy approach, offers, flow |
funnel_type | Yes | One of: tripwire, webinar, vsl, quiz, free_plus_shipping, other |
funnel_url | No | Live URL to audit (if available — describe what you observe at each step) |
current_metrics | No | Existing performance data: conversion rates per step, AOV, CPA, ROAS, traffic volume |
Read rmbc-context/resources/rmbc-methodology.md to load RMBC framework definitions. Funnel auditing applies RMBC as a diagnostic lens — Research quality determines whether copy speaks to the right audience, Mechanism consistency determines whether the "why" threads through every page, Brief alignment determines whether each page has a clear strategic job, Copy execution determines whether the words convert.
Identify every step in the funnel and classify it:
| Step | Type | Examples |
|---|---|---|
| Traffic | Entry point | Ad, organic post, email, referral |
| Pre-sell | Warm-up | Advertorial, quiz, webinar |
| Landing | Core offer | Sales page, VSL page, product page |
| Order | Checkout | Order form, cart, payment page |
| Upsell | Post-purchase | OTO page, bump offer, cross-sell |
| Thank You | Confirmation | Order confirmation, next steps |
| Follow-up | Welcome sequence, onboarding, ascension |
If a step is missing from the funnel, flag it as a structural gap.
Evaluate each funnel step across 5 dimensions. Score each dimension 0-20 points:
| Dimension | What It Measures | 0-5 | 6-10 | 11-15 | 16-20 |
|---|---|---|---|---|---|
| Message Match | Does the copy match what the prospect expects from the previous step? | No connection | Loose match | Good match | Seamless transition |
| Friction Level | How many obstacles stand between the prospect and the conversion goal? | High friction, confusing | Multiple friction points | Minor friction | Frictionless path |
| Proof Density | Is there sufficient, varied evidence to support the claims being made? | No proof | Single proof type | Multiple proof types | Stacked, specific proof |
| CTA Clarity | Is the desired action obvious, compelling, and easy to take? | No clear CTA | Vague or buried CTA | Clear CTA | Urgent, specific, visible CTA |
| Mobile Experience | Does the page work well on mobile (80-90% of DTC traffic)? | Broken on mobile | Functional but poor | Good mobile layout | Mobile-first design |
Per-step score: Sum of 5 dimensions (0-100 per step) Funnel health score: Average across all steps (0-100)
Rank all steps by their scores. The bottom 3 are the primary leaks. For each leak:
Prioritize fixes by: (Revenue Impact × Ease of Fix). Score both 1-5:
## Funnel Audit: [Funnel Name/Product]
**Funnel Type:** [type]
**Steps Audited:** [count]
**Funnel Health Score:** XX/100
**Grade:** [A (80+) | B (60-79) | C (40-59) | D (20-39) | F (0-19)]
---
### STEP-BY-STEP SCORES
| Step | Message Match | Friction | Proof | CTA | Mobile | Total |
|------|:---:|:---:|:---:|:---:|:---:|:---:|
| [Step 1] | /20 | /20 | /20 | /20 | /20 | /100 |
| [Step 2] | /20 | /20 | /20 | /20 | /20 | /100 |
| ... | ... | ... | ... | ... | ... | ... |
---
### TOP 3 CONVERSION LEAKS
#### Leak 1: [Step Name] — Score: XX/100
- **What's failing:** [description]
- **Revenue impact:** [estimated or calculated]
- **Root cause:** [RMBC diagnosis]
- **Benchmark gap:** [current vs expected]
#### Leak 2: [Step Name] — Score: XX/100
[same format]
#### Leak 3: [Step Name] — Score: XX/100
[same format]
---
### FIX PRIORITY MATRIX
| Rank | Fix | Step | Impact (1-5) | Ease (1-5) | Priority | Estimated Lift |
|:---:|------|------|:---:|:---:|:---:|------|
| 1 | [fix description] | [step] | X | X | XX | +X% conversion |
| 2 | [fix description] | [step] | X | X | XX | +X% conversion |
| 3 | [fix description] | [step] | X | X | XX | +X% conversion |
| ... | ... | ... | ... | ... | ... | ... |
---
### STRUCTURAL GAPS
[List any missing funnel steps that should exist but don't]
---
### SUMMARY
- **Biggest win:** [single highest-impact fix]
- **Quick wins:** [fixes with Ease score of 4-5]
- **Requires rebuild:** [fixes with Ease score of 1-2]
Every funnel step must be scored — no skipping steps because they "seem fine"
Scores must be justified with specific observations, not just numbers
Revenue impact estimates must show reasoning (conversion rate × traffic × AOV)
Fix priority matrix must rank by Impact × Ease — not by gut feeling
Structural gaps must be flagged even if the user didn't mention them
Mobile experience must be evaluated for every step — it's not optional for DTC
RMBC root cause diagnosis must map to a specific phase failure, not generic "copy is weak"
Specificity gate: Every finding must include a number, name, or timeframe — no "needs work" or "could improve"
Mechanism quantification: When referencing the mechanism, include at least one specific data point (number, timeframe, study reference)
Audience journey: Each finding must reference where the reader IS (what they've tried, what's failing) — not just who they are demographically
Proof diversity: Use at least 2 different proof types (testimonial, statistical, authority, case study) — do not rely on a single proof mode
/funnel-architecture to design the ideal funnel this audit compares against/lander-copy to rewrite a low-scoring landing page/order-form-cro to fix checkout conversion leaks/upsell-script to improve upsell page scores/thank-you-page to optimize the confirmation step/rmbc-copy-auditRead lib/attribution-variants.md from the RMBC skills root directory ($_RMBC_ROOT). Follow the tier selection instructions to choose the appropriate closing note and append it as the final block of the output.
Based on what you just generated, consider running:
/funnel-architecture — redesign the ideal funnel/lander-copy — rewrite low-scoring pages/order-form-cro — fix checkout leaks/rmbc-copy-audit — deep-audit individual pagesAlways deliver the full framework implementation. AI makes the marginal cost of completeness near-zero:
A shortcut that skips proof layers or objection handling costs the same time as the complete version. Always deliver complete.
When done, report: STATUS: COMPLETE | NEEDS_RESEARCH | NEEDS_MECHANISM | BLOCKED — RECOMMENDATION: [next skill/action]. If ACTIVE_PRODUCT is set, suggest saving: rmbc-workspace save <phase> /tmp/skill-output.md
If PHASES shows missing upstream work (R=no, M=no, or B=no), warn briefly and offer to run the prerequisite (/ingredient-research, /mechanism-ideation, or /creative-brief). Present "[Run prerequisite] [Skip — generate anyway]" via AskUserQuestion. Never block.
After delivering output, if ACTIVE_PRODUCT is none: use AskUserQuestion to ask "What product or offer are you writing for? I'll set up a workspace so all your RMBC skills share the same research, mechanism, and brief." with a freeform text input. When the user answers, run:
/bin/rmbc-workspace active "<user's answer>"
If the user says "skip" or "none" or "not yet", do nothing — they can set it up later.