Enterprise Future Guide: Capabilities become a unit of investment and accountability

Koko: Welcome to Koko Knows: Enterprise Future Guide. I'm Koko, and joining me are Sam and Max. Today's episode is a standalone expansion of one topic from the parent audio Enterprise Future Guide: Capabilities, leadership and value in 2030 and 2035. The claim: by 2030, the capability may become a real unit of investment—not a function, not a workflow, but something with an owner, economics and a measurable outcome.

Max: Max here. Before we go further—what distinguishes a capability from a function in budget terms? Because from an operator's seat, if I'm still funding finance, ops and commercial separately, the word capability can just be a new label on the same structure.

Koko: That's the precise test. A capability like fulfill a profitable customer promise spans commercial, planning, operations and finance. It needs one economic baseline, one outcome owner and a measurable decision gate. Without those, it's a coordination idea, not a unit of accountability.

Sam: I'm Sam, and I want to flag the evidence ceiling here. McKinsey and BCG both describe this direction, but those are advisory frameworks built on early examples. BCG's cases include projected returns, not necessarily realized ones. That's not a reason to dismiss the direction, but it limits the confidence we should assign.

Koko: Agreed. The NBER Cybernetic Teammate study—a preregistered field experiment with 776 P&G professionals—found AI helped bridge commercial and technical thinking in bounded innovation tasks. One organization, bounded scope. It's directional signal, not evidence of enterprise redesign at scale.

Sam: And Bain's finding is the sharpest counterweight: AI-focused reorganizations showed weaker reported performance than other reorgs. The explanation is difficulty changing everyday work, not technology. That's exactly where a capability layer can add management cost without removing the original handoffs.

Max: So what changes the hypothesis? What would tell us this is working versus adding a matrix on top of a matrix?

Koko: Outcome improvement after full coordination, control and transition costs—compared with a credible, well-run functional alternative. Fewer recurring exceptions. Faster resolution of cross-functional conflicts with a traceable owner. If those measures don't move, narrow the change.

Sam: And professional accountability has to stay structurally distinct. Finance integrity, legal authority, independent assurance—those can't be absorbed by a capability owner. The owner coordinates commitments across those professionals; the owner doesn't replace them.

Max: That's the governance point that matters for 2035 as much as today. If agents take on more execution, the question of who holds authority and who provides independent challenge becomes more consequential, not less.

Koko: So the action is concrete: a business owner, COO and CFO select one customer promise—order-to-cash is a reasonable candidate—map the handoffs, establish an economic baseline and test explicit outcome ownership. The decision gate is measurable improvement in margin, delivery reliability or cash conversion before touching any reporting line.

Sam: Test the work and the incentives first. Retain a stop or redesign option. That's the discipline Bain's finding argues for.

Koko: For the deeper operating contract—what a capability needs in terms of data, authority, assurance and portability—the companion reading is titled Enterprise capabilities in this series. Thanks for listening to Koko Knows: Enterprise Future Guide.