CMO Forcing Questions

This skill runs a narrative-first interrogation of your marketing plan, forcing you to answer six critical questions about ICP, JTBD, positioning, distribution, CAC payback, and brand defensibility. It produces a structured CMO review with a verdict (green/yellow/red) and next steps.

✨ What it does

  • Forces naming one real ICP person, not a persona
  • Checks JTBD with a one-sentence customer statement
  • Validates one-sentence positioning statement
  • Identifies primary distribution channel and inbound/outbound intent
  • Evaluates CAC payback per channel (flags >18 months as hobby)
  • Assesses brand defensibility against copycat competitors

🎯 When to use it

  • Before launching any new campaign
  • Before changing positioning, tagline, or category
  • Before allocating more than 10% of marketing budget to a new channel
  • Before a major PR moment (funding announcement, product launch)
  • When pipeline contribution is declining

🚀 How to use

Trigger: /cs:cmo-review

Trigger with the slash command /cs:cmo-review <plan> where <plan> is your marketing plan or description. The skill expects you to answer six forcing questions in writing, optionally run two Python models, and then apply the verdict. Example prompts:

/cs:cmo-review Our new SaaS product launch plan targeting mid-market CFOs.
/cs:cmo-review We're repositioning our brand and need a strategic review.

📄 Output: A markdown CMO review report with positioning statement, ICP details, channel mix, verdict, and next steps.

📦 Add this skill to Claude Code

# 1. Get the skills repo
git clone --depth 1 https://github.com/alirezarezvani/claude-skills /tmp/claude-skills

# 2. Copy this skill into your project (or ~/.claude/skills for all projects)
mkdir -p .claude/skills
cp -r /tmp/claude-skills/c-level-agents/skills/cmo-review .claude/skills/cmo-review

Skill source: c-level-agents/skills/cmo-review/SKILL.md

⚠️ Good to know

Requires you to provide a plan and answer the six questions; the skill does not automatically gather data or run the models unless you execute the commands.

❓ FAQ

What does the verdict mean?

Green means the story is sharp and channel mix sound; yellow means sharpen positioning before scaling; red means positioning is broken and you should not spend.

What if I can't name a real ICP person?

The skill says if you can't name one, the ICP isn't sharp enough, implying you need to refine your ICP before proceeding.

What is considered a valid alternative in JTBD?

The skill states that 'We use spreadsheets' is a valid alternative, and so is 'we don't.'

🤖 Overview, features, install steps and FAQ were generated from the project's SKILL.md on Sep 4, 2026. Always check the original source before running commands.