The Seasonal Planning Protocol
Overview
A planning framework for the high-velocity AI era where traditional annual or rigid roadmaps fail. It replaces fixed timelines with "Seasons" defined by secular industry changes.
Core principle: Build for the slope, not the snapshot.
The Framework
┌─────────────────────────────────────────────────────────────────┐
│ 1. DEFINE THE SEASON │
│ Based on secular trend (e.g., "Season of Agents") │
│ Not calendar-based │
├─────────────────────────────────────────────────────────────────┤
│ 2. SET LOOSE QUARTERLY OKRS │
│ Directionalize, don't lock down │
│ Allow for capability changes │
├─────────────────────────────────────────────────────────────────┤
│ 3. EXECUTE 4-6 WEEK SQUAD SPRINTS │
│ Tight execution loops │
│ Concrete deliverables │
├─────────────────────────────────────────────────────────────────┤
│ 4. LEAVE SLACK FOR DISRUPTION │
│ Don't fill capacity to 100% │
│ Room for new model capabilities │
└─────────────────────────────────────────────────────────────────┘Key Principles
| Principle | Description |
|---|---|
| Season-based | Secular changes define seasons, not calendar |
| Loose OKRs | Direction, not destination |
| Short sprints | 4-6 weeks, concrete outcomes |
| Built-in slack | Reserve capacity for the unexpected |
Common Mistakes
- Adhering to rigid annual plans
- Filling capacity to 100% without innovation buffer
- Treating all quarters the same despite industry shifts
*Source: Asha Sharma (Microsoft AI Platform VP) via Lenny's Podcast*