Okta SVP Offers SQUAD Framework To Measure True AI Business Value
The Question
How should engineering teams measure AI ROI beyond token costs?
The Lesson in This Case
Okta SVP of Engineering Monica Bajaj argues that token consumption tracks spend rather than outcomes, and proposes the SQUAD framework, covering Security, Quality, Unit Economics, Acceleration, and Developer Experience, to fix that. The framework shifts measurement toward metrics such as cost per merged pull request, vulnerability catch rates before code merges, and PR review cycle time, while dynamic model routing directs simple tasks to lighter models rather than expensive frontier ones. Senior leaders evaluating AI budgets should take seriously Bajaj's warning that siloed engineering, product, and design teams can push net AI ROI negative by generating low-quality output that requires costly rework downstream.
Originally reported by Forbes. Read the full story here.