Microsoft VP: SaaS becomes AI agent layer, not a 'toggle tax'

The piece starts from a gap between AI pilots and transformation. Teams report research and drafting that once took months now taking days, yet momentum slows because one part of the workflow has not changed: people still bridge the gaps between systems. For decades enterprise software has been built around that manual stitching, so the app, not the data underneath, has been what people engage with.

The argument is that agents change that equation. If agents can reach the same data, the question shifts from why keep the app to what the app becomes. The author's answer is that SaaS supplies the context, relationships, governance, business rules and logic that agents need to act reliably and securely, which is why the "SaaSpocalypse" reading misses the point.

What remains open is execution. The body pairs the promise with a measure of organizational readiness: only one in four employees see leadership alignment on AI, and the productivity J-curve means gains arrive after an organization aligns on a vision and redesigns the work. Whether the toggle tax actually shrinks at scale is likely to hinge on that redesign, and the people most affected are the operations, finance and HR staff whose day is currently spent moving information between applications.