ADMF — Adaptive Digital Marketing Framework

A 15-step, AI-guided marketing system that takes a product from raw documentation to a fully optimized, AI-visible, data-driven marketing operation.

What ADMF Does

ADMF is a collection of structured prompts that guide an AI agent (Claude, GPT, Gemini, etc.) through building a complete marketing stack for any product. You point it at your product documentation, and it walks you through:

Phase 1: Foundation — Parse your docs, inventory features, build personas, map the competitive landscape Phase 2: Strategy — Core messaging, pricing narrative, channel strategy Phase 3: Content Engine — Production copy, PAS story templates, media generation (video, ads) Phase 4: Optimization — Analytics feedback loop that updates the stack based on real performance data Phase 5: AI Visibility — AEO setup (llms.txt, structured data), deep research whitepapers, LLM influence monitoring Phase 6: Distribution — Marketplace listings (G2, Reddit, ProductHunt), community engagement playbooks

What makes it different

• Interactive facilitator — choose-your-own-adventure navigation, not a rigid sequence
• Content flywheel — Steps 9→10→11→9 create a continuous production + optimization loop
• AI visibility — Steps 12-14 make your content discoverable by ChatGPT, Perplexity, Claude, and Google AI Overviews
• Project-agnostic — works with any product. You provide the docs, ADMF provides the framework

Quick Start

1. Copy the framework into your project

# Clone this repo
git clone https://github.com/rossres/admf-org.git

# Copy core/ into your project
cp -r admf-org/core/ your-project/_admf/core/

2. Create your config

cp your-project/_admf/core/config-template.yaml your-project/_admf/config.yaml

3. Fill in config.yaml

At minimum, set:

• project_name — your product name
• user_name — your name
• source_documents — paths to your product documentation (PRD, brief, pitch deck, README — anything that describes what you're building)

Leave brand_voice, banned_words, and visual_system empty — the framework establishes these during Steps 1/5/8.

4. Create the output directory

mkdir -p your-project/_admf/output

5. Run it

Tell your AI agent: "Run ADMF" or "Build marketing stack"

The facilitator will greet you, show your progress, and let you pick where to start.

The 15 Steps

Phase	Step	Name	What it produces
Foundation	1	Session Setup	Parsed docs, buyer context, initialized output
Foundation	2	Feature Inventory	Stack-ranked features with customer-facing value
Foundation	3	User Personas	Buyer profiles with pain points, triggers, language
Foundation	4	Competitive Positioning	Matrix, category definition, "only we" statements
Strategy	5	Messaging Framework	Core narrative, taglines, vertical hooks
Strategy	6	Pricing Narrative	Pricing story, value anchors, objection handling
Strategy	7	Channel Strategy	GTM plays, vertical priorities, content calendar
Content Engine	8	Content Generator	Complete copy deck, image prompts, asset manifest
Content Engine	9	PAS Story Templates	Problem-Agitation-Solution narratives per persona
Content Engine	10	Media Production	Video, static ads, search ads via user-selected tools
Optimization	11	Analytics Feedback	Performance data → diagnosis → stack updates
AI Visibility	12	AEO & AI Readiness	llms.txt, schema markup, AI-optimized content
AI Visibility	13	Deep Research	Consumer question mining → authority whitepapers
AI Visibility	14	LLM Influence Discovery	Monitor what shapes AI answers, find citation gaps
Distribution	15	Marketplace & Community	G2, Reddit, ProductHunt listings + engagement

How It Works

The facilitator (core/facilitator.md) is the entry point. When invoked, it:

1. Loads your config.yaml
2. Checks for existing sessions
3. Shows a journey map with your progress
4. Lets you jump to any step (with dependency warnings)

You don't have to run steps in order. The facilitator warns you about dependencies but never blocks you.

File Structure

your-project/
  └── _admf/
      ├── core/                    ← Framework (don't edit — update from repo)
      │   ├── facilitator.md       ← Interactive agent guide
      │   ├── template.md          ← Output document skeleton
      │   ├── config-template.yaml ← Blank config for new projects
      │   └── steps/               ← The 15 step files
      │       ├── step-01-session-setup.md
      │       ├── ...
      │       └── step-15-marketplace-community.md
      ├── config.yaml              ← YOUR project config (edit this)
      └── output/                  ← YOUR generated content
          ├── marketing-stack-{date}.md
          ├── images/
          ├── videos/
          ├── whitepapers/
          └── ...

Requirements

• An AI agent that can read markdown files and follow structured instructions (Claude Code, Cursor, Windsurf, Gemini CLI, etc.)
• Product documentation (any format — PRD, brief, pitch deck, README)
• For Steps 12-14: a live website with analytics
• For Step 15: accounts on target platforms

Examples

See examples/saas-example-config.yaml for a complete config example showing a fictional SaaS product.

License

MIT