It is the last working week before the holidays, and the finance team is planning the year-end close: five working days to close the books, with the auditors arriving the Monday after. Nadia sets the calendar, everyone gets an owner and a deadline, and one number is already known to be late.
OK, let's start. Year-end close: five working days, and the auditors arrive the Monday after. So this year the calendar has no spare day in it.
Nadia does not open with an agenda. She opens with the one fact that shapes every decision in the meeting: there is no spare day. Starting with the constraint tells people why the rules that follow are strict.
Last year it had two, and we used both.
Graham is saying "last year we were late" without blaming anyone. Senior people in finance often make a criticism as a plain fact with no adjective. Listen for what is not said.
We did. So I've built this one differently. Every task has a day, an owner and a definition of done. I'll go through the days first, then the owners, then the things that worry me.
"First the days, then the owners, then the things that worry me." A short spoken map helps everyone follow, and it is especially useful when you are chairing in a second language: you have told people where you are going, so nobody is lost if you pause.
Good. Start with the days.
Day one: the ledgers close. Purchase ledger and sales ledger, both at noon. Anything that arrives after noon belongs to January.
This is cut-off: deciding which period a transaction falls into. "Belongs to" is the natural verb. You will also hear "falls into January" and "goes into the next period".
Sorry, can I check one thing? When you say "close", do you mean nothing new gets posted, or that everything posted has been reviewed?
"When you say X, do you mean A or B?" is one of the most useful questions in any meeting. Aisha does not say "I don't understand". She offers two possible meanings and asks which one is right. It sounds precise, not junior.
Good question. Nothing new gets posted. Review is day three. They're different words on purpose.
OK. Posted by day one, reviewed by day three. Got it.
Aisha repeats the answer in her own short version. This confirms she understood and gives Nadia a chance to correct her. It takes five seconds and prevents a day-three problem.
Day two: accruals and prepayments. Day three: intercompany agreed, and all balance sheet reconciliations done. Day four: a first full set of numbers goes to Graham. Day five is review, adjustments and sign-off.
And day five is mine. I'd like to spend it reading, not chasing.
To chase someone is to ask them again and again for something that is late. Graham is setting an expectation politely: everything should reach him on day four without him having to ask.
Then the owners. Aisha, you have the bank reconciliations, and the supplier statement reconciliations for our top twenty suppliers.
Top twenty by spend?
By spend, yes. That covers about eighty per cent of what we buy.
I've already asked all twenty for a statement dated the thirty-first. Fourteen have confirmed. I'll call the other six before the break.
What I did, where it stands, what I will do next. Aisha gives a complete status in three short sentences and nobody needs to ask a follow-up question. This is the shape of a good update in any sector.
Good. That's the kind of thing that saves us a day in January.
And when is a reconciliation done? When the two balances agree?
When every difference has a name next to it. The balances don't have to agree. I just need to know why they don't.
Nadia corrects a common idea: a reconciliation is not finished when the numbers match, but when every difference is explained. "Has a name next to it" is an informal way to say "is identified".
And if one of them doesn't agree to our ledger?
In accounting, two records "agree" when they show the same balance. You will hear "agree to the ledger", "tie to the statement" and "reconcile to the bank". If they do not agree, there is a "difference" — not an "error" until someone has explained it.
Then you don't fix it. You flag it. Write down the supplier, the amount and what you think it is, and bring it to me the same day.
To flag something is to make it visible to the right person. Nadia tells Aisha exactly what a good flag contains: who, how much, and your best guess. A flag with a guess is far more useful than "there's a problem".
Even if it's small?
Especially if it's small and you can't explain it. A small difference you understand is fine. A small difference nobody understands is the one that grows.
Nadia turns Aisha's question around: "Even if it's small?" — "Especially if it's small." This is a strong, natural way to correct an assumption without saying "no, you're wrong".
She's right. I once signed off a set of accounts with a four-hundred-pound question in them. It was not four hundred pounds by March.
Graham backs Nadia in front of the team, then gives a short personal example where he was the one who made the mistake. A senior person admitting an old error makes it safer for a junior person to report a new one.
Matteo, you have the accruals pack: bonus, freight and supplier rebates. I have intercompany, because Rotterdam needs chasing, and they answer me faster than they answer anyone else.
They answer you because they are afraid of you.
Whatever works.
But I need to raise something now, and not on day two. The rebate accrual. I can't give you a final number on day two.
"I need to raise something now" is a clear signal that a problem is coming. Matteo names the item first, then states the limit plainly: "I can't give you a final number." No apology, no long introduction.
Go on.
Our rebates depend on what we bought in the full year. The big suppliers send their December volume statements on day three, sometimes day four. So on day two I have eleven months of actuals and one month of my own estimate.
"Actuals" are real, recorded figures. An "estimate" is a calculated guess. Finance people keep these two words strictly apart, and Matteo is careful to say exactly how much of his number is each.
How big is the number, and how far can it move?
Graham does not ask why it is late. He asks for size and range. If you bring a risk to a senior finance person, prepare these two answers before you walk in.
The accrual is about one point two million for the year. December could move it by sixty thousand, up or down.
Sixty either way. That's not nothing, but it's manageable. What's your estimate based on?
A very British way to weigh a problem: it is real, but it is not serious. Then Graham immediately asks for the basis. In finance, a number without a basis is only an opinion.
Our own purchase orders for December, at the rebate rates in the contracts. The only thing I can't see is whether we crossed the next volume tier with two of the suppliers.
Matteo separates what he knows from what he does not know, and he keeps the unknown small and specific. This builds trust: people believe your estimate more when you tell them exactly where it is weak.
So here's what I'd suggest. On day two you post the estimate, and you label it as an estimate, with the basis written in the journal. On day four, when the statements are in, you post the true-up as a separate journal.
Nadia is the one who decides, but she frames the decision as a suggestion. "I'd suggest" is softer than "you will", and in practice everyone understands it is the plan. A "true-up" is the adjustment that brings an estimate to the real figure.
Two journals. Not one corrected one.
Two. I want the auditors to see what we knew on day two and what changed. If you overwrite it, it looks like we guessed and got lucky.
Nadia explains why. People follow a rule more carefully when they understand what it protects. Here the reason is the audit trail: a visible record of what was known, and when.
And on day four, will I see both numbers?
You'll see three, on one line. The estimate, the true-up, and the difference between them.
Then I'm comfortable. Go ahead on that basis.
In finance, "I'm comfortable with that" is approval. It means "I have seen enough to accept this". The opposite, "I'm not comfortable", is a polite but serious objection. "On that basis" ties the approval to the conditions just agreed.
Fine. An estimate with a label on it. I can do that.
And Matteo, thank you for saying it today. That is the rule I want for the whole week. No surprises.
Graham thanks Matteo in front of everyone before he states the rule. He is showing the team what happens when you bring a problem early: you are thanked, not blamed.
Can I ask what counts as a surprise? I mean, things will go wrong.
Another definition question from Aisha. "What counts as X?" asks where the line is. It is a polite way to test a rule that sounds simple but is not.
Of course they will. Bad news on day two is information. The same news on day five is a surprise. The only difference is when I hear it.
Graham's rule in one contrast. Notice that he does not ask for no problems. He asks for early problems. This sentence returns later in the series.
So if you suspect something, tell Nadia the day you suspect it. Don't wait until you're sure. Being wrong early costs us nothing.
Understood. Suspect early.
Two working rules, and then we're done. First, the close tracker. One shared file, every task, four statuses: not started, in progress, done, blocked. Update it when the status changes, not at the end of the day.
And "blocked" means what, exactly? Because last year everything of mine was "in progress" for three days.
Matteo laughs at himself, but the question is serious: a status that is always "in progress" tells nobody anything. A small joke against yourself is a safe way to raise a weakness in last year's process.
I remember. Blocked means you're waiting for someone else and you can't move. If you've been blocked for more than half a day, it comes to me.
This is an escalation rule, said very simply. Nadia defines the word, sets a time limit, and names herself as the person who takes the problem. Compare it with the weaker "let me know if there are any issues".
Half a day. OK. That's strict.
It's five days, Matteo. Half a day is ten per cent of the close.
Matteo says the rule is strict. Nadia does not argue or soften it. She gives one number that makes the rule look reasonable. In finance, arithmetic is often the most polite argument.
Second, late invoices. After the noon cut-off, anything under five thousand pounds goes into January and we don't discuss it. Anything over five thousand, we accrue it, and it goes on Matteo's list.
A threshold is the amount above which a different rule applies. "Anything under… anything over…" is the standard way to state one. "We don't discuss it" means the decision is already made, so nobody loses time on small items.
My list is getting long.
Is five thousand the same threshold as last year?
The same, and it was agreed with the auditors at the planning meeting. I didn't want to change a threshold in the year they're looking at cut-off.
Graham asks a yes-or-no question. Nadia answers "the same", then adds the two things he really wants to know: the auditors agreed it, and she chose not to change it for a reason.
Good. Leave it.
One more from me. Intercompany. Last year Rotterdam and the UK disagreed until the last evening. What's different this time?
A fair and direct challenge. Graham does not say "don't let it happen again". He asks what has changed in the plan, which is a question Nadia can actually answer.
I've asked Rotterdam for their balances on day two, a day before I need them. And Warsaw has agreed to the same. If there's a difference, I'd rather have a whole day to find it.
And they've said yes?
Warsaw said yes. Rotterdam said they'll try, which is why I'm chasing them myself.
Nadia hears the difference between a commitment and a hope. "We'll try" is not a yes, and she plans around it instead of trusting it. Listen for this in your own meetings: who said yes, and who said they would try?
So, to say it back. Day one, ledgers close at noon. Day two, accruals from Matteo, with the rebate estimate labelled. Day three, intercompany from me and reconciliations from Aisha. Day four, numbers to Graham. Day five, sign-off.
Day, task, owner — five times, with no extra words. A recap like this is the most valuable thirty seconds of a planning meeting. If someone disagrees with their line, this is the moment they will say so.
And the Monday after, the auditors.
One request for day four. With every number that moved against last year, I want one sentence saying why.
One sentence?
One. If it needs two, come and see me before you send it.
When a figure changes from one period to the next, finance people expect a short explanation with it. This is often called commentary. Graham's limit of one sentence forces people to find the main reason and leave out the rest.
One last question. If I find something on day three, in a reconciliation, and Nadia is in the middle of Rotterdam, do I wait?
Aisha turns the general rule into a specific situation she can imagine. Asking "what do I do if…?" before the busy week is much better than guessing during it.
No. You interrupt me. Rotterdam can wait. An unexplained difference can't.
Many junior people, and many non-native speakers, wait too long because interrupting feels rude. Nadia removes the doubt with a direct instruction. If your manager has not told you this, it is worth asking.
That's the meeting. Enjoy the break, all of you. I'll see you on day one.
With eleven months of actuals and a lot of hope.
Label it.
Graham answers Matteo's joke with Nadia's rule from earlier: label the estimate. Ending on a shared reference shows the team has already absorbed the rule, and it sends everyone out in a good mood.