Lisa's departure left a critical execution gap. Content quality, volume, and speed have been persistent issues. Planned roadmap items from Q4 2025 remain unbuilt. Meanwhile, buyer intelligence from 841 call transcripts shows we've been telling our story instead of the story buyers need to hear. The Content Machine addresses all of this simultaneously.
An AI-powered content system that flows from intelligence to published assets, with Claude handling the heavy production work and the team focused on strategy and quality assurance.
Notion is the brain, the single source of truth for all content planning, assets, context, and workflows.
Claude is the engine, generating briefs, drafting content, running brand QA, repurposing assets, monitoring competitors, and producing performance reports.
The team is quality assurance, reviewing, approving, providing strategic direction, and making the final call on what ships.
This inverts the traditional model. Instead of a person doing 80% of the work and AI assisting with 20%, the machine does 70% and the team focuses on the 30% that requires human judgment.
Every piece of content flows through 7 stages with two human checkpoints. Everything between checkpoints is automated.
| Stage | What Happens | Who |
|---|---|---|
| 💡 Idea | Idea Engine scans market signals, threat intel, customer data, and competitive moves, seeds pipeline with prioritized ideas | AI |
| ✅ Approved | Team reviews ideas in Notion, approves the ones worth pursuing | Human |
| ⚙️ In Production | Campaign Manager plans the campaign → Content Executor produces all assets → Brand QA runs automatically | AI |
| 📄 Drafts Ready | Anchor + derivative drafts staged as nested pages in Notion with QA results | AI |
| 🔍 In QA | Team reviews drafts, provides editorial feedback | Human |
| ✔️ QA Approved | Content cleared for publishing | Human |
| 📢 Published | Team publishes to channels (blog, LinkedIn, email, etc.) | Human |
The system runs across three layers, each with a distinct role.
| Task | Schedule | What It Does |
|---|---|---|
| Idea Engine | Mon / Wed / Fri, 8 AM | Scans market + intel → seeds Content Pipeline with ideas |
| Pipeline Watcher | 4x daily, weekdays | Monitors for Approved items → triggers Campaign Manager + Executor |
| Competitive Monitor | Weekly (Monday) | Scans competitor activity → updates Competitive Landscape database |
| Weekly Big Rocks | Monday, 7 AM | Generates Brian's 6 weekly priorities → posts to Marketing Team Meeting page |
20 specialized Claude skills organized in 5 categories, each trained on the company's brand voice, terminology, buyer personas, and competitive landscape.
Idea Engine calls → Pulse of the Market + Intel Gathering + Customer Research Intel + Competitive Intel → synthesizes 3-7 ideas → seeds Content Pipeline
Campaign Manager reads approved idea → selects anchor asset type → maps 4-6 derivative pieces across channels → assigns Writer Skills
Content Executor reads campaign plan → routes each asset to its Writer Skill → runs Brand Terminology QA on every piece → stages all drafts in Notion under the anchor item
The machine handles volume and consistency. The team handles judgment and direction.
| Component | Status |
|---|---|
| Content Pipeline (Notion database + Board view) | ✅ Live |
| Idea Engine (skill + scheduled task) | ✅ Live |
| Campaign Manager (skill + pipeline watcher) | ✅ Live |
| Content Executor (skill) | ✅ Live |
| All 4 Writer Skills | ✅ Live |
| Brand Terminology QA | ✅ Live |
| Competitive Monitor (scheduled) | ✅ Live |
| Weekly Big Rocks (scheduled) | ✅ Live |
| Clickhouse MCP (Gong, ChurnZero, Pendo) | ✅ Connected |
| Enhancement | Priority |
|---|---|
| Slack notifications when drafts hit "Drafts Ready" | High, pending channel confirmation |
| Verify Clickhouse intel-gathering with real data | High |
| Hootsuite integration for automated social posting | Medium, future phase |
| GA / Pardot data for Monthly Performance Digest | Medium, future phase |
| Salesforce MCP for attribution tracking | Low, future phase |