---
id: "question-measuring-flywheel-velocity"
type: "open-question"
source_timestamps: ["§ Type 4: Data Flywheels and Lock-In Ecosystems"]
tags: ["metrics", "data-science"]
related: ["concept-data-flywheels"]
resolutionPath: "Publishing industry-specific case studies detailing the exact mathematical KPIs used by data science teams to track model improvement per data ingestion cycle."
sources: ["spine"]
sourceVaultSlug: "hbr-seg-spine"
originDay: 1
articleStem: "hbr-edu-47-5-types-ai-investment"
sourceUrl: "https://hbr.org/2026/06/the-5-types-of-ai-investment-and-how-to-capture-their-value"
sourceTitle: "The 5 Types of AI Investment–and How to Capture Their Value"
---
# What are the specific formulas for measuring 'flywheel velocity'?

**Open question.** For [[concept-data-flywheels|Type 4]] investments, the financial logic depends on measuring *flywheel velocity* — how fast the AI improves per cycle of operational data. The article gives the conceptual definition but leaves open the **mathematical or operational formulas** needed to track this velocity across different industries (e.g., agriculture vs. software).

**Resolution path.** Publishing industry-specific case studies detailing the exact mathematical KPIs that data-science teams use to track model improvement per data-ingestion cycle. This gap makes it hard to compare flywheel strength across a portfolio and is the practical bottleneck for acting on [[action-invest-closed-loop-systems]].