What a CFO should ask before approving a new exception

Every exception a CFO approves — a non-standard term, a one-off discount, an unusual write-off — takes a moment to grant and can take hours to explain years later if nobody thought to ask the right questions at the time. The approval itself is rarely the bottleneck. What determines whether the exception ages well is a short list of questions worth asking before signing off, not after something's already been questioned.

Why the approval moment is the only moment that matters

Once an exception is approved, it typically runs unexamined for a long time — that's the nature of an exception, as distinct from something reviewed on a recurring basis. The approval moment is therefore the only reliable point at which the full context is guaranteed to be available, because the CFO approving it is, by definition, looking closely at the situation right then. Every day after that, the context degrades — see institutional memory has a shelf life — while the exception itself keeps running exactly as approved.

The questions worth asking every time

What specifically makes this case different from the standard policy? If the answer is vague or generic, that's worth noticing before approval, not after — a genuine exception usually has a clear, specific answer to this question.

Is this a one-off or the start of a pattern? An exception granted once, in isolation, is a different kind of decision than an exception that will implicitly become the new normal for similar future requests. Being explicit about which one this is prevents an accidental precedent from being set without anyone intending it.

What would need to change for this to no longer be appropriate? This single question does more work than almost any other, because it converts the exception from a static, permanent-feeling arrangement into something with a built-in condition for review — see the exception that made sense once and confuses everyone now.

Who else needs to know this exists, and where will they find out? An exception that only the approver knows about is invisible to anyone else who might reasonably need to account for it later — a successor, an auditor, a colleague managing a related account.

Why this takes minutes and prevents hours

None of these questions require extensive analysis — they're quick to ask and quick to answer, precisely because the CFO approving the exception already has the relevant context fully in mind. The value isn't in the depth of the questions. It's in the fact that they're asked at all, at the one moment the answers are cheap to produce, rather than left to be reconstructed later at much greater cost — see reconstructing a write-off decision two years later.

Turning this into a habit rather than a checklist nobody follows

Attach the questions to the approval step itself, as a required part of granting any exception above a defined threshold, rather than relying on the CFO to remember to ask them informally each time.

Keep the answers with the exception, not in a separate log. The value of the answers depends entirely on someone being able to find them later, attached to the decision they explain — see where to write down why so it survives.

Review the accumulated exceptions periodically, using the recorded review triggers as the starting point. If every exception was approved with a stated condition for revisiting it, the periodic review becomes a simple check against those stated conditions, rather than an open-ended investigation into every exception from scratch.

Why this specifically belongs at the CFO level

Exceptions significant enough to require CFO approval are, almost by definition, the ones with the highest downside if their reasoning is later lost — larger amounts, more scrutiny, more consequence if questioned. Building this discipline in at the point of highest-stakes approval protects the exceptions most worth protecting, without requiring the same rigour to be applied to every minor, low-risk decision across the business.

Common questions

Why is the moment of approving an exception the best time to capture its context?

Because that's the only point at which the full reasoning is guaranteed to be readily available — the approver is actively considering the situation right then. After approval, exceptions typically run unexamined for long periods, and the context degrades the longer it sits uncaptured.

What's the most valuable question to ask before approving an exception?

What would need to change for this to no longer be appropriate. This question converts a static, seemingly permanent arrangement into something with a built-in trigger for future review, which prevents the exception from running indefinitely on inertia rather than continued justification.

Why does this discipline matter more at the CFO approval level than elsewhere?

Because exceptions significant enough to require CFO sign-off tend to be the ones with the highest downside if the reasoning behind them is later lost — larger financial impact, more scrutiny from auditors or the board, and more consequence if the decision is ever questioned.

How can a business make this a consistent habit rather than an occasional good practice?

Build the questions into the approval workflow itself as a required step for exceptions above a defined threshold, and keep the recorded answers attached directly to the exception rather than in a separate log, so they're actually findable when a future review or question requires them.

Related: institutional memory has a shelf life · reconstructing a write-off decision two years later · where to write down why so it survives