Lab
Experiments
Every format starts as a hypothesis. Test it, measure it, then scale it or kill it.
Sample data
How to use this screen
- 01Write a hypothesis for a new format (e.g. 'Cold-open hooks lift retention').
- 02Run it on a small batch, record the result, then scale it or kill it.
Live today: Experiments listed are examples of how to write them.
- What goes in
- Ideas from founders and patterns from analytics.
- What the system does
- Tracks each experiment from idea → planned → testing → result.
- What you decide
- Whether an experiment succeeded and should become standard.
- Where it goes next
- Winning formats become rules for clipping and content.
idea 1planned 1testing 1successful 1failed 0scaled 0
- EXP-001Successful
Obey Every Traffic Law
Self-imposed restraint in a chaos game creates natural tension and payoff.
- Owner
- Jarrod Parkhurst
- Project
- GTA VI Content Engine
- Created
- 1 Oct
- Confidence
- 82%
- Result
- Completion 71% vs 48% baseline across 5 posts.
NextScale to a weekly series.
- EXP-002Testing
Start a Taxi Company
Long-running business arcs build return viewers.
- Owner
- Elliot
- Project
- GTA VI Content Engine
- Created
- 1 Oct
- Confidence
- 40%
- Result
- —
NextPublish episodes 1–3, compare follow rate.
- EXP-003Planned
NPC Interviews
Treating NPCs as interview subjects yields reliable comedy hooks.
- Owner
- Elliot
- Project
- GTA VI Content Engine
- Created
- 1 Oct
- Confidence
- —
- Result
- —
NextWrite ten interview prompts.
- EXP-004Idea
Attempt to Run a Legitimate Business
Earnest effort inside a criminal world is inherently funny.
- Owner
- Jarrod Parkhurst
- Project
- GTA VI Content Engine
- Created
- 1 Oct
- Confidence
- —
- Result
- —
NextDecide on business type.
Statuses: Idea → Planned → Testing → Successful → Failed → Scaled