The Seasonal Planning Protocol
by Asha Sharma • Corporate Vice President of Product, AI Platform at Microsoft
Former COO at Instacart, VP of Product at Meta (Messenger/Instagram Direct), Early employee at Porch Group.
🎙️ Episode Context
Asha Sharma discusses the paradigm shift from building products as static artifacts to living organisms in the age of AI. She covers the rise of agentic workflows, the necessity of 'post-training' investment, and how Microsoft structures product planning amidst rapid technological change.
Problem It Solves
The inability to plan accurate long-term roadmaps when underlying technology (like GPT models) changes rapidly.
Framework Overview
A planning framework designed for the high-velocity AI era where traditional annual or rigid roadmaps fail. It replaces fixed timelines with 'Seasons' defined by secular industry changes, allowing for agility while maintaining a north star.
⚡ Step-by-Step Framework
Identify the 'Season' based on secular changes (e.g., 'Season of Agents').
Set loose quarterly OKRs to directionalize the team.
Execute in 4-6 week squad-based goals.
Deliberately leave 'slack' in the system for the unplanned and the slope of innovation.
Identify the 'Season' based on secular changes (e.g., 'Season of Agents').
Set loose quarterly OKRs to directionalize the team.
Execute in 4-6 week squad-based goals.
Deliberately leave 'slack' in the system for the unplanned and the slope of innovation.
When to Use
In hyper-growth or rapidly evolving industries (like AI) where 6-month plans become obsolete quickly.
Common Mistakes
Adhering to rigid annual plans; filling capacity to 100% without leaving room for new model capabilities.
Real World Example
Microsoft defining a 'Season of Prototyping' followed by a 'Season of Agents' to align thousands of engineers.
We think about it as what season are we in? ... I feel like you have to actually build for the slope instead of the snapshot.
— Asha Sharma