What an LP’s operational diligence team is actually testing

{{TODO: CLIENT INPUT}} - body to be written. Outline below, generated from the build spec. Delete this note and set status: published when the piece is written.

Outline

Opening. The DDQ asks about governance in language that sounds procedural. It is not procedural. Almost every question in the operations section is a proxy for one thing: can any single person do something material without a second pair of eyes.

The leading veto trigger. Thin independent oversight, and specifically weak segregation of duties around cash and valuation. Cite F5 through the Fact component. Explain why this outranks strategy and track record in the assessor's ranking, which is that a control failure is unbounded and a bad quarter is not.

Why small firms fail this by default. Six people cannot separate every incompatible duty by headcount. This is a real constraint and it is worth saying so plainly rather than pretending otherwise. The answer is not more people. It is a permission model that makes the separation structural, plus compensating controls where headcount genuinely does not allow it.

What "compensating control" means here. Concrete examples: dual release on payments above a threshold, reconciliation by someone with no initiation rights, an independent review of related-party transactions, an approval chain that is configuration rather than convention.

What the assessor is looking at in the file. Not the policy. The evidence that the policy operated: approval records with two distinct names, a reconciliation with a preparer and a reviewer, an access review with a date.

The valuation half. The same test applied to marks. Method, inputs, approver, change history. Why an audit trail assembled after the request is worth less than one produced as the work happened, and how an assessor can tell the difference.

Close. The distinction between having controls and being able to demonstrate them is the entire subject. Link to the readiness practice.