Day three of the close, two o'clock. Nadia and Matteo take the credit-note problem to Graham. The difference is real, it is bigger than materiality, and nobody knows yet whether last year has the same problem. Nadia has to give him the facts, the size, two options and a recommendation, in that order.
Graham, thanks for making time. We've found a problem in the close. It isn't a disaster, but it isn't small, and I'd rather you heard it today.
Nadia gives the size of the problem in one sentence before any detail: not a disaster, not small. A senior listener can then calibrate everything that follows. Opening with background makes them guess how worried to be.
Who else knows?
Graham's first question is not about the numbers. He wants to know who might already be telling a different version of the story, and who might be surprised. In an escalation, controlling who hears what, and when, comes before the detail.
The three of us, Aisha, who found it, and the payables lead, because I asked her for a list. Nobody outside finance.
Good. Then go on. Start with what happened, not what it means.
Graham sets the order: what happened, then what it means. Facts first keeps opinion out of the picture until everyone agrees on the facts. It also makes the meeting shorter.
Payables record supplier documents by the date they receive them, not the date on them. So credit notes dated December that arrived after Christmas went into January. There are eleven of them.
Eleven. From how many suppliers?
Seven. The biggest is Ferrand, a hundred and forty-eight thousand. In total, two hundred and eighty-six.
How sure are you of the eleven?
Sure of the list. Payables pulled it from the system, and Aisha checked it against three supplier statements. All three matched.
Graham asks how sure, and Nadia answers with where the number came from and how it was checked. "Sure of the list" is precise: she is sure of what was found, not yet of what it means.
And how did we find it?
Aisha's reconciliation for Ferrand didn't agree. She flagged it this morning, with a guess, the way we asked.
Nadia names the person who found the problem, in front of the director. It costs nothing, it is true, and it shows that the process from the kickoff worked.
Right. Now what it means. Matteo, I assume the numbers are yours.
They are. Of the two hundred and eighty-six, a hundred and seventy are rebate credit notes. Most of those were already in my accrual as estimates. The only real change there is Ferrand. We crossed tier two, so fifty-two thousand more than I estimated.
Matteo does not give one total and stop. He splits the two hundred and eighty-six into the part that was already expected and the part that is new. Breaking a number down is how you show you understand it.
And the second supplier you weren't sure about?
I called them at noon. No tier credit. We didn't cross it. So my estimate for them stands.
At the kickoff, Matteo had two unknowns. Here he closes the second one in four short sentences: what he did, what he heard, what it means, what changes. A good escalation also reports the things that turned out fine.
And the rest? A hundred and sixteen.
Returns and price corrections. Nobody accrued those, because nobody knew they were coming. So our December costs are overstated by the full hundred and sixteen.
Why so many returns in December?
Our customers over-ordered in November, before Christmas, and sent stock back in December. It happens every year. This year the returns were simply bigger, so the credit notes from our suppliers were bigger too.
Graham asks why, and Matteo answers in three short steps: what customers did, that it is normal, and what was different this year. A one-line cause is often all a director needs before moving on.
Are all seven suppliers ours, or are some of them Rotterdam's?
Two are Rotterdam's. Thirty-one thousand between them. I haven't told Rotterdam yet. I wanted you to hear it first.
Tell them today, straight after this. They'll need to correct their side before day four.
Nadia held the news back from Rotterdam so that her director would not hear it second-hand. Graham then sets the next step and the deadline. The order in which people hear bad news is part of handling it.
So, fifty-two plus a hundred and sixteen. A hundred and sixty-eight thousand of profit that belongs in December and isn't there.
Graham adds the two parts himself, slowly, out loud. Senior people often check the sum in front of you. It is not distrust; it is how they make the number their own before they take a decision on it.
Yes. And materiality is a hundred and fifty. So we're above it.
Materiality is the size at which an error matters to someone reading the accounts. Nadia puts the number next to the line the auditors drew. That one comparison tells Graham that leaving it is not an option.
Above it, but not by much. And in the direction nobody complains about.
Except the auditors.
Except the auditors. Fair.
Graham repeats Nadia's three words and adds one. It is a short, natural way to accept a correction from someone junior to you, with no loss of authority.
How long have payables been doing this?
Since the new purchasing system went live. Two years ago.
So last year's close could have the same problem.
It could. I don't know yet. I'd rather say that than guess.
This is one of the hardest sentences to say to a director, and one of the most useful. Nadia adds "yet", which signals that finding out is already planned, and "I'd rather say that than guess", which explains why she isn't offering a number.
When will you know?
Not this week. Checking last January means pulling last year's documents, and Aisha has six reconciliations still to do. I'd like to do it the week after the auditors start, and tell Alan that it's coming.
Nadia separates this year's close from last year's question. She gives a date for the second one, and a reason why it can't be now. Saying "not this week" with a reason is better than promising something you can't deliver.
And to be fair, if it happened last year too, part of it would cancel out. Last year's late credits would have landed in this year.
Part of it. Not all of it. I don't want "it probably cancels out" written anywhere.
Matteo's point is technically true. Graham accepts the logic but refuses the conclusion, because "probably" is not evidence. In finance, an argument that makes a problem smaller needs the same proof as one that makes it bigger.
One more question for you, Matteo. Does tier two change anything for next year?
A little. If we stay at Ferrand's volume, next year's rebate is about sixty thousand higher than the budget assumes. I'd rather raise it at the budget review than change the budget today.
Agreed. Write it down for the budget review, with the basis.
Matteo spots a second effect but chooses the right meeting for it. Graham agrees and adds one condition: write it down, with the basis. Parking a point is fine only if someone owns it and it is recorded.
Then options. There are two I'd consider.
One: move all eleven credit notes into December, whatever the size. Two: move the nine that are over five thousand pounds, and leave the two small ones in January, under the threshold we agreed with the auditors.
"One: … Two: …" Numbering options out loud makes them easy to compare and easy to refer to later: "I'd go with two." Two options are usually enough; a third is often a straw man.
What's the difference in money?
Seven thousand. The two small ones together.
And your recommendation?
Option two. It's consistent with the threshold we already use for late invoices. If we treat credit notes differently from invoices, Alan will ask why, and I won't have a good answer.
Nadia gives her choice and one strong reason: consistency. Then she says what would happen otherwise. One good reason is more convincing than four average ones.
Before I decide, argue option one for me. Why would someone choose it?
It's simpler to explain. Every December credit note is in December, no exceptions. If Alan prefers no exceptions, it's already done.
And you still prefer two?
I still prefer two. The threshold exists so that we don't spend day four on seven thousand pounds.
Graham asks Nadia to make the case for the option she did not choose. If she can argue it fairly and still prefer her own, her recommendation is stronger. It is a common test in finance, and a good habit before any decision.
And it's quicker. Nine journals instead of eleven isn't much, but tomorrow is day four.
I don't care about two journals. I care about the answer to Alan's question. But I agree with the reason. Option two.
Graham accepts the option but sets aside Matteo's argument for it. He is telling the team which reasons count. A decision made for the right reason is easier to defend later.
And the process. If we fix the eleven and change nothing else, we'll be here again in March.
From Monday, payables record by the date on the document. And in the first week of every quarter, I review every credit note over five thousand myself. I'll own both, and I'll have them in place before the first-quarter close.
Nadia proposes a process change, a check that catches it if it fails, an owner and a deadline. This is the shape of a "management response" that auditors expect, and it will come back later in the series.
Could the system stop it? A warning when the date on a document falls in a closed month?
Good idea. I'll ask IT what it costs. If it's cheap, it could replace my manual check.
Ask, but don't wait for it. The manual check starts this quarter, whatever IT says.
Matteo suggests an automatic control. Graham welcomes it but keeps the manual check as the plan. A good idea with no date should not delay a fix that already has one.
Last thing. The auditors start on Monday. Who tells Alan?
I'd like to tell him myself, first thing on Monday, before he opens a single file.
Why before?
Because if he finds it in a sample, it's his finding. If we tell him first, it's our correction.
Nadia explains why the timing matters. The facts will be identical, but who reports them first changes how they are recorded and how much the auditor trusts the rest of the numbers. Notice the parallel structure: "it's his finding … it's our correction."
Agreed. And I'll call his partner on Friday afternoon, so it isn't news to anyone at their end either.
Graham applies his own rule to the auditors. He will tell Alan's senior partner before Monday, so that nobody on their side hears it second-hand. The rule from the kickoff works upward and outward, not only inside the team.
What exactly will you say to him? The first two sentences.
We found a cut-off problem with credit notes in this year's close, and we've corrected it. It may also have existed last year, and we'll check and send you the result by the twentieth.
Good. Write those two sentences down, and say them exactly like that. Don't add a third.
Graham asks for the exact opening, not a summary of it. For a sensitive conversation, preparing the first two sentences word for word is one of the most useful things you can do. Under pressure, people add explanations; "don't add a third" stops that.
And my rebate true-up? I've held it, like Nadia asked.
Post it tomorrow, with Ferrand at tier two, as a separate journal. And Matteo, your day-two estimate was labelled and it had a basis. That's the only reason we can see the gap so clearly.
Matteo's estimate was wrong by fifty-two thousand. Graham praises it anyway, for the right thing: it was labelled and had a basis. Praising good habits when the outcome is imperfect is how teams keep those habits.
I'll take that. It's better than I expected from this meeting.
And by the end of day four, I'd like all of this on one page. The eleven, the journals, the process change, and your two sentences for Alan.
One page. By the end of day four.
Nadia repeats the request in four words. Short confirmations like this show you heard the exact deliverable and the deadline, and they close the item.
Nadia, you came in with the problem, the size, two options and a recommendation, on day three. That's exactly what I asked for. It's not a nice afternoon, but it's the right one.
Graham lists exactly what Nadia brought. Specific praise teaches the whole team what a good escalation looks like: problem, size, options, recommendation, early.
Thank you. And Aisha?
I'll tell her myself. On my way out.