Handover documents can describe the normal process and leave out why a founder set it aside. Before the handover ends, I would have the acquiring team decide one past exception from the record while the founder watches and stays quiet.
Take a hypothetical industrial supplies distributor bought by a larger competitor. Its order system puts any account more than 60 days past due on credit hold and blocks shipment. That rule is configured and easy to hand over. The founder ran credit, and the system log shows them releasing holds on the same handful of accounts several times a year. Those releases are the exception worth testing, because the rule cannot say which holds to lift and the log does not say why.
Choose one recent release that looks like the others, with real money on the order. Rebuild the file as it stood that morning: the aging report, the blocked order, the account notes, and the credit team’s email with that customer. Strip out everything that happened afterward. Hand it to the person on the acquiring team who will own credit and ask for a decision with reasons: ship, hold, ship part, or call someone first. The founder answers nothing during the exercise, and the credit lead writes down every question they would have asked.
A live hold would be a more realistic test and a worse idea. If the team gets it wrong, a good customer’s shipment stops or product goes to an account about to default, and the company pays for the exam. A wrong call on a past case leaves current orders alone. Its weakness is that the team knows it is being tested, so a good result is softer evidence than it looks.
Suppose the credit lead holds the order. The founder then explains the release. Part of the overdue balance was disputed over a short shipment, confirmed by the warehouse in an email the credit team never saw. This customer also pays in one batch at quarter close, which the founder knew from years of watching it happen and never recorded. That gives the remaining handover two specific jobs: a way to get the warehouse’s confirmation into the credit file, and a written note on that account’s payment timing.
The exercise tests whether this file held enough for this decision. A clean result says little about allocating stock during a shortage or approving a price concession, where the founder may keep context somewhere else. A wrong call is weak evidence about the credit lead’s judgment, since the exercise was built to expose missing inputs. The founder’s past decision is also one data point. It may have been right for reasons that no longer apply, or lucky. If the credit lead would still hold after hearing the full story and can say why, that is a policy choice the new owner is entitled to make.
I would spend what remains of the handover on the credit lead’s written questions, each answered inside the system the team will use, with the founder checking the wording. Then run a second case from a different kind of exception. No number of cases proves the transfer is complete, and the departure date will arrive either way. For any account still carrying an open question on that date, I would require a second approval on releases until the buyer has watched that customer through a few payment cycles of its own.