COLLABORATION MODELS
Model Types
| Model |
Description |
Best For |
| AI-First |
AI drafts, human refines |
High-volume content |
| Human-First |
Human directs, AI assists |
Strategy, creative |
| Parallel |
Both work, then merge |
Research, ideation |
| Sequential |
Handoffs between AI/human |
Complex workflows |
Model Selection Guide
| Task Type |
Recommended Model |
| Research/data gathering |
AI-First |
| First drafts |
AI-First |
| Creative direction |
Human-First |
| Strategy development |
Human-First |
| Ideation |
Parallel |
| Production content |
Sequential |
ROLE DEFINITIONS
AI Role
| Task |
AI Does |
AI Doesn't |
| Research |
Gather, summarize |
Validate, decide |
| Writing |
Draft, variations |
Final approval |
| Analysis |
Process, pattern-find |
Interpret meaning |
| Creative |
Generate options |
Select direction |
Human Role
| Task |
Human Does |
Human Doesn't |
| Strategy |
Decide, direct |
Manual data gathering |
| Quality |
Review, approve |
Every first draft |
| Creative |
Direction, judgment |
Repetitive variations |
| Customer |
Empathy, nuance |
Routine responses |
HANDOFF POINTS
Standard Workflow
Human: Define brief/requirements
↓
AI: Research and gather information
↓
Human: Review and validate research
↓
AI: Generate outline/first draft
↓
Human: Edit, enhance, approve
↓
AI: Format, optimize
↓
Human: Final review and publish
Handoff Checklist
| From |
To |
Handoff Includes |
| Human → AI |
Clear brief, context |
Requirements, examples |
| AI → Human |
Draft, sources |
Confidence notes |
| Human → AI |
Feedback |
Specific edits |
| AI → Human |
Final version |
Change summary |
QUALITY CHECKPOINTS
Checkpoint Matrix
| Stage |
Check |
Owner |
| After research |
Accuracy, completeness |
Human |
| After draft |
Brand, accuracy, value |
Human |
| After edit |
Quality score |
Human |
| Before publish |
Final approval |
Human |
Quality Gate Criteria
| Gate |
Criteria |
Threshold |
| Research |
Sources verified |
100% |
| Draft |
Quality score |
35/50 |
| Final |
Quality score |
40/50 |
| Publish |
Human approval |
Required |
FEEDBACK LOOPS
Feedback Types
| Type |
Purpose |
Frequency |
| Prompt feedback |
Improve AI outputs |
Every use |
| Quality feedback |
Track quality trends |
Weekly |
| Process feedback |
Optimize workflow |
Monthly |
| Tool feedback |
Improve tools |
Quarterly |
Feedback Template
Prompt used: [paste]
Output quality: [1-10]
What worked: [describe]
What didn't: [describe]
Suggested improvement: [describe]
CONTINUOUS IMPROVEMENT
Improvement Cycle
- Track AI usage and quality
- Identify common issues
- Update prompts/processes
- Share learnings
- Measure improvement
Metrics to Track
| Metric |
Frequency |
Owner |
| Time savings |
Monthly |
Team lead |
| Quality scores |
Weekly |
Individual |
| Error rate |
Monthly |
Team lead |
| User satisfaction |
Quarterly |
Manager |