The Decision-Native Enterprise

Prologue — The threshold

This book was written at a threshold. On one side of it, machine intelligence was a tool that individuals used to work faster. On the other, it is a factor of production that enterprises must organize around, the way they once organized around steam, electricity and information systems. Most companies are still standing on the first side, buying licenses, running pilots, measuring adoption, while the ground under the org chart has already moved. The technology is ahead of the institutions. This book is about the institutions.

It makes one central claim: the organization you run was designed as a solution to a problem that no longer exists. Information was scarce, so hierarchies were built to move it. Coordination was expensive, so meetings, reports and approval chains were built to manufacture it. Cognition was the bottleneck, so work was divided into fragments a single human mind could hold. Every one of those constraints has now either fallen or is falling, and an organization that keeps the structures without the constraints is paying rent on a building it no longer occupies. The question this book will teach you to ask of everything in your enterprise, every meeting, every role, every approval, every system, is the question archaeologists ask of ruins: what constraint was this built for? Ask it honestly and your organization stops looking inevitable and starts looking historical, which is the beginning of being able to change it.

To be clear about the target before a single wall comes down: the demolition eyes this book gives you are for scaffolding, not for craft. The production sequence, the engineering method, the clinical protocol, the specialist depth of your functions — the things that actually execute value — are not the ruins. They are what the ruins grew around, and the book will spend as much care teaching you to protect them as to clear what encases them.

A word about the most seductive wrong turn available at this threshold, because you will be offered it this quarter if you have not been already. Process intelligence, the mining and modeling of your workflows in exquisite detail, is being sold as the transformation's first step, and as the final stage of the dashboard era it is a magnificent achievement. It is also a trap in exactly the way peak versions of old paradigms are traps: the fastest sailing ships were built after the steamship existed. Perfect visibility into your processes becomes custody of them; custody becomes optimization; and the organization ends up polishing its scaffolding to the decimal while believing itself transformed. The escape is not to refuse the tool but to reassign it: treat the mined map as a demolition survey, the clearest picture ever taken of what the audit in Chapter 4 should question first.

How to read it. Part I is the argument, and it is built to be read in one sitting, tonight. Board members and executives can stop after Part I and Chapter 8 and know exactly what they are being asked to decide. Operators and program leads should live in Part III, which is a field guide, and photocopy Chapter 10. The skeptical, the burned and the risk-minded should start at Chapter 17, the failures, and Chapter 18, the fights, and work backwards; the book was written to survive that reading order. And everyone gets Part VI, the toolkit, which is meant to be stolen.

The undone is the opportunity, and whoever waits inherits second place.