A synthesis of more than 30 years of restructuring research by Sam Rees reached a consistent finding: most organisational restructures fail to deliver their stated objectives. The current wave of AI-driven restructures shows no evidence of being different.

The research identifies the usual culprits: change management fatigue, unclear ownership of redesigned processes, and a tendency to reorganise structure without redesigning the work.

What is specific to the AI moment is the expectation of speed. Boards and executive teams are under pressure to demonstrate AI adoption in quarters, not years. That pressure is producing structural changes that are decoupled from the operating model changes required to make them work.

McKinsey's case study this week on "redesigning how software gets built" when AI enters the workflow is instructive. The organisations that saw productivity gains were not the ones that deployed AI tools the fastest. They were the ones who rebuilt the team structures and decision rights around the new capabilities - a slower, more deliberate process that looks like under-performance in the short term.

The Zoom webinar trend is also telling. "How to Handle Employee Resistance to AI" is now a mainstream category for corporate events. In most cases, resistance is not the problem. Resistance is information about implementation quality.

Organisations that treat it as an obstacle to manage rather than a signal to interpret will achieve compliance without adoption. The restructure wave is real. The question is whether the organisations driving it are prepared to do the slow work.

https://lnkd.in/gt88PJWp

#OrganisationalDesign #AITransformation #ChangeManagement