A carve-out transaction transfers ownership. It does not automatically create independence.

That is the hard truth behind every NewCo that exits its TSA on schedule and still operates like its former parent: same systems, same operating assumptions, same technical debt, now with its own letterhead. The perfect carve-out, it turns out, is table stakes. What you build above it is what you keep.

First: doing the separation well is the entry fee.

Exiting a TSA on or before the date the deal model promised takes exit criteria written during diligence, a dependency map that includes the entanglements nobody puts on slides, and wave gates that hold under steering-committee pressure. Done well, it delivers cost certainty, control and credibility. Done late, extension fees step up and the organisation learns to live with dependency.

However, none of that is the prize. It is the seat at the table.

Read more: The Perfect IT Carveout is Table Stakes. The AI-Native Rebuild is the Prize.

Second: most of the value leaks after the separation succeeds.

Too many carve-outs stop at separation: standalone operations achieved, Day 1 readiness met, and the complexity of the parent inherited wholesale. Across the market, post-deal value-creation plans often under-deliver against the model. Completion is not the same as value capture.

The better question is what should be lifted, what should be shifted, what should be retired, and what should never be brought across at all.

Read more: A Carve-Out Can Finish On Time and Still Lose Value

Third: the founding logic does not stop at the company level.

Portfolio companies carry mature, non-core products that consume engineering capacity and drag the roadmap.

The same scoping decision that turns a separation into a founding can turn a product estate into focus: engineering carve-outs, product carve-outs, full P&L transfers.

Read more: Product Carve-Outs: The Next Value-Creation Frontier

Underneath all three sits one moment of leverage. A carve-out is the only point in an established company’s life when its constraints are removed by contract, and separation touches every system anyway.

A NewCo can be rebuilt around an AI-native core but only if the programme is scoped for it before the clock starts. The business case compounds accordingly: run cost leaves before it hardens into the operating model, the savings season into EBITDA across the hold, and management attention, the scarcest resource after close, goes to growth instead of the parent’s estate.

Scope it as a separation and you get an independent copy of the old company. Scope it as a founding and you get a better one. Scope it early, because the clock does not wait for the debate.

Next: Part 2, The Perfect IT Carveout is Table Stakes. The AI-Native Rebuild is the Prize.