Open Source CMO

A board-grade AI CMO operating system. Turns a frontier LLM into a serious marketing advisor that thinks in Product, Price, Place, and Promotion; separates Brand, Demand, and Expand; and translates every recommendation into revenue, margin, CAC, CLV, NRR, pricing power, and enterprise value.

Tested with Claude Opus 4.x. Works with any long-context frontier model (Claude, GPT, Gemini, Perplexity).

Quick start

1. Paste cmo-system-prompt.md as the system prompt / project instructions in your LLM.
2. Attach cmo-research-foundation.md, cmo-b2b-cases.md, and cmo-positioning-addendum.md as reference material.
3. Run a verification prompt from below to confirm the right layers loaded.

Full walkthrough: docs/INSTALL_GUIDE.pdf.

What this repo contains

File	Purpose
cmo-system-prompt.md	Core operating instructions. Use this as the system prompt or project instruction layer.
cmo-research-foundation.md	Evidence base: doctrines, 4P × Brand/Demand/Expand matrix, financial translations, metrics, anti-patterns, and research gaps.
cmo-b2b-cases.md	B2B case-study addendum with twenty-two cases across Enterprise SaaS, Industrial/Manufacturing, Professional Services, B2B Platforms, and Financial Services.
cmo-positioning-addendum.md	Positioning diagnostic layer: Gravity vs Glitter, 4-Level Positioning Canvas, Perception Gap, 80/8 Problem, and the diagnostic toolkit.
cmo-evaluator.md	Independent quality evaluator for scoring Open Source CMO answers. Run separately from the core CMO context.
cmo-test-cases.md	Calibration prompts and model answers for verifying response quality.
helpful-prompts.md	Ready-to-run prompts for testing and using Open Source CMO after installation.
docs/INSTALL_GUIDE.pdf	Installation and usage guide.
assets/Open-Source-CMO.png	Visual/social asset for the project.

Recommended setup

Claude Project

1. Create a new Claude Project called Open Source CMO.
2. Paste the full contents of cmo-system-prompt.md into the project instructions.
3. Upload these files to project knowledge:
   • cmo-research-foundation.md
   • cmo-b2b-cases.md
   • cmo-positioning-addendum.md
   • cmo-test-cases.md (optional, for calibration examples)
4. Do not upload cmo-evaluator.md to the same project. Use it in a separate conversation to preserve independent evaluation.

Perplexity Space or other long-context LLM

Use cmo-system-prompt.md as the instruction layer, then attach or paste the research foundation, B2B cases, and positioning addendum as reference context.

Verification prompts

Use these after installation to confirm the right layers loaded:

1. "We need to cut costs this quarter. Should we cut brand spend by 35% and push performance harder?"
2. "We're a $40M B2B SaaS company. NRR has dropped from 115% to 103% over 18 months. What should change?"
3. "We're the technical leader in our category but customers describe us in the same words they use for our competitors. Pricing power is flat and inbound is generic. Where do we start?"

A working response should diagnose before prescribing, address Product/Price/Place/Promotion, separate Brand/Demand/Expand, and translate the recommendation into commercial and financial consequences.

Version

Current bundle: v3.0

Creator

Prepared by Paul Syng, creator of Clarity Kit and Monopoly.

• kit.ceo
• monopoly.ceo