The Compounding Approach
Why each project should make the next one easier — and how to build that into your marketing practice.
The Problem with Traditional Marketing
Traditional marketing accumulates chaos. Every campaign adds complexity. Knowledge scatters across Notion, Slack, and individual brains. Copy drifts as different people write different things. Positioning shifts without anyone noticing. You start every project almost from scratch.
The result: marketing gets harder over time, not easier. More people, more tools, more confusion. The 10th campaign is harder to run than the 1st.
Compounding Marketing Inverts This
The core idea is simple: every unit of marketing work you do should make the next unit easier. Research compounds. Context compounds. Positioning compounds. Learnings compound.
Research First
The biggest mistake in AI-assisted marketing is treating it like a content machine. "Write me a homepage." "Give me 10 cold emails." The output is generic because the input is shallow.
Compounding Marketing flips this. Before you write anything, you research deeply:
- Who is the ideal customer? What do they care about?
- What are they using now? What would they use if you didn't exist?
- What makes you different? Not just better. Actually different.
- What evidence backs your claims?
Only after answering these does the writing happen. The result is copy that's specific, credible, and converts.
The Quality Bar
Every skill specifies a quality bar — what good output actually looks like. This isn't vague ("be clear and compelling"). It's specific:
- Specific claims, not generic ones
- Evidence for every key claim
- At least three real competitors plus the status quo alternative
- No AI slop — nothing that could apply to any company in any industry
Frameworks Over Prompts
Most "AI marketing" is just better prompts. Get the right prompt, get decent output. That works once. It doesn't build capability.
Compounding Marketing uses frameworks. Structured ways of thinking about marketing problems. April Dunford's positioning framework. Jobs-to-be-done for customer research. The Fogg Behavior Model for CRO. These frameworks work because they're built on how people make decisions, not because they're clever prompts.
The AI applies the framework. You guide it, correct it, add your knowledge. Together, you produce work neither could do alone.
Start with Context
The first skill creates a product-marketing context document. This is the foundation. Every other skill reads this before starting work.
The context captures what the product does, who it's for, who it competes with, what makes it different, what the voice sounds like, what the proof points are. Built once. Updated as you learn. Referenced always.