12 January 2026
A 13-week cash view is a habit, not a spreadsheet
A cash model updated once and left in Downloads is a drawing. The value sits in the Friday refresh and in the conversation it forces that same afternoon.
Mid-sized groups in Singapore often discover cash in the week, not in the monthly pack. A large receipt slips. Payroll hits. The managing director opens the bank and feels the week. Meanwhile the finance team has a closing-balance forecast built after the last scare and left untouched. The file is not the problem. The absence of a rhythm is the problem. Thirteen weeks is long enough to see a squeeze coming and short enough that people will still tell the truth about invoices they actually expect.
The model itself can be plain. Receipts from the sales ledger, with collection days as a named driver rather than a hidden plug. Payroll, rent, tax and known supplier runs on the weeks they actually leave the account. A line for the items you know are late. If the books and the bank disagree, that gap is visible. A tidy closing balance achieved by a plug is a report that will fail the first week someone is on leave, because nobody else knows which cell is the fiction.
What makes the habit hold is a one-page snapshot and a named owner. The snapshot goes into a standing meeting. The owner updates it on a Friday from the same extracts, in a file the company owns. Assumptions sit on one sheet so a change in collection days can be typed while people are still in the room. We have watched groups recover two or three weeks of warning simply by putting the large receipts on a list with a date.
What to try this month. Name the person who will update a thirteen-week view this Friday. Put collection days on a single sheet. Write down the three receipts you are counting on, with dates.
23 January 2026
Why your gross margin changes depending on who is asked
Finance and the commercial lead can each be right about margin and still talk past one another. The missing object is a shared unit written down.
Ask a site manager, a finance lead and a salesperson for gross margin and you may receive three percentages, all internally consistent, none of them usable in the same meeting. One is excluding delivery. One is after a discount that lives only in the CRM. One is a monthly blend. The quarrel is rarely about arithmetic. It is about the unit and about which costs are allowed into the room. Until those two things are written down, every dashboard will look like an argument with colours.
We start by naming the unit the business actually sells. A chair-hour in a clinic. A job in freight. A site-day in F&B. A cohort of logos in B2B software. Then we list the costs that belong next to that unit. Delivery, success-team time, payment fees, waste, overtime: they are either in or they are out, and the dictionary says which. A cost that is “sometimes included, depending who built the file” is how a group ends up with two truths.
The second step is a reconciliation to the P&L. An operational margin that cannot be walked to the ledger will be waved away in the first ten minutes of a board meeting, and rightly. The walk can be approximate at first. It cannot be absent. Where a cost is still estimated, we label it as estimated and we give it an owner. Hiding the estimate in a blended rate is how it survives into the next quarter.
What to try this month. Write the unit on one line. List the five costs that may sit next to it. Ask finance and operations to tick the same list. The cells they disagree on are the beginning of the dictionary.
6 February 2026
Chart of accounts: the cheapest reporting fix there is
Before a warehouse and a dashboard, many groups need a chart a human can still read. That unglamorous work pays back every month.
We are often asked to put the numbers in a BI tool when the ledger still holds three versions of rent, a miscellaneous bucket the size of a small department, and product revenue that cannot be cut the way the commercial meeting speaks. A dashboard on top of that chart will faithfully display the mess, faster. The cheapest reporting improvement is still a chart of accounts that matches how the business is run, with a mapping for the history so last year does not fall apart.
This is a short list of decisions. Which revenue lines must stand alone because pricing lives there. Which cost lines must stand alone because a manager is held to them. Which historic codes can be mapped and retired rather than kept alive out of superstition. A bookkeeper who closes every month already knows most of this. The work is to write it down, agree it with the finance lead, and then stop inventing new codes in the heat of a close.
While the chart is being tidied, reporting can still start. We build a mapping table and we run the first pack through it. That is slower than a perfect chart, and it is honest. Groups that wait for a clean year before they build a pack often wait through another year of hallway numbers. Groups that map as they go, and then fold the mapping into the books at a quiet period, get a pack and a better ledger.
What to try this month. Export the chart. Highlight every code used fewer than a handful of times last year, and every code whose name you cannot explain to a new hire. Those two lists are the agenda for a one-hour meeting with whoever closes the books.
19 February 2026
Board packs that survive the first question
A pack fails when the chair’s first question cannot be answered from the page. Design those pages around the questions the room actually asks.
We have sat in rooms where a thirty-page pack was sent on Tuesday and the meeting on Thursday still began with whether the figure was the same as Xero. The pages were full. The questions were not. A board in a mid-sized group usually wants a small set of decisions: cash, margin mix, hiring, pricing, and whether a banking conversation is still comfortable. Each of those can live on a page with a figure, a short note on what moved, and a pointer to the detail.
Language matters. If the chair says “contribution after delivery” and the pack says “gross profit”, someone will spend four minutes translating, and the decision will wait. We write the pack with the finance lead, in the words the room already uses, and we put those words in the dictionary so the dashboard does not drift. Commentary is structured. It is not a novel, and it is not a paste from last month with the dates changed. If nothing moved, the page can say so in one sentence. Silence is better than theatre.
Reconciliation is the unglamorous page that saves the meeting. A one-line walk from the management view to the close, with a named threshold for the residual, tells the room they are looking at one set of books. Without it, every number is optional. With it, the chair can ask a real question about mix or cash and expect an answer that still holds after lunch.
What to try this month. Write the five questions you believe the chair will ask. Next to each, write the figure you would put on the page, and whether you can walk it to the close today. The gaps are the build list.
2 March 2026
Rolling forecasts for owner-managed groups
An annual budget still argued in April is already late. A rolling view keeps the conversation on what the group now believes this month.
Owner-managed groups in Singapore often have a budget that was serious work in December and a relic by the time hiring, rent or a large customer has moved. The annual file is not useless. It is the plan you said you would run. What the group needs beside it is a forecast that can change when the facts change, without pretending the plan never existed. Variance then has a home: we planned this, we now believe that, here is why.
Drivers make the refresh possible. Headcount, volume, price, rent, a currency on a material, a collection-day assumption: they sit on a sheet with names. Actuals land beside them. The forecast is the plan restated with what you now know. Scenarios are copies with labels, “delay the hire”, “the rate moves”, not a nest of files in a personal drive. The quarterly review is a dated meeting with a pack, not a weekend of archaeology.
We still meet groups that treat a forecast as a second budget they must defend, which is how the numbers stop moving. A forecast that cannot be wrong in public will be wrong in private. The discipline is to write the assumption, date it, and let the next month’s actuals talk back. That is slightly uncomfortable for a room that likes a single annual truth. It is also how a PE-backed board, or a bank conversation, stays adult.
What to try this month. Keep the annual budget frozen as the plan. Build a second view, even a rough one, with three drivers you can update from last month’s actuals. Take that view to one operating meeting and see whether the conversation changes. If it does not, the drivers are probably still the wrong ones.
Tell us what decision is stuck
If a note is uncomfortably close to the file on your desktop, request a scoping call. We reply within one working day, Singapore hours.