Processes
Monthly financial close without the side spreadsheet
In many companies, the first week of the month is lost. Someone in finance exports data from three systems, pastes it into a spreadsheet, checks it line by line and, around the tenth, delivers a result that was already out of date. By the time the number arrives, the next month is half over.
Why the close takes so long
The close itself is simple: check what came in, what went out and what is still pending. What eats the time is gathering the information, which is scattered. The most common reasons:
- Sales in one system, billing in another, the bank in a third. None of them talks to the others, and the connection is made by hand.
- Entries with no category or the wrong one. They only show up at closing time, when nobody remembers what they were.
- The side spreadsheet. Each area keeps its own control "because the system's version does not match". At month-end there are three versions of the same number.
- Checking concentrated in one person. When they go on vacation, the close runs late or does not happen at all.
The side spreadsheet deserves special attention. It starts as a solution: someone needed a view the system did not provide. Over time it becomes an area's official source, and the company ends up with more than one truth.
The cost that does not show
The damage of a slow close is not just finance's time. It shows up in three places:
- Late decisions. Pricing, hiring and investment are decided based on the month before last.
- Errors that repeat. A duplicate charge or a wrongly entered discount is only noticed weeks later.
- Distrust. When two areas' numbers do not match, the meeting turns into an argument about which spreadsheet is right.
Gartner estimated, in 2020 research, that poor-quality data costs the surveyed organizations an average of US$12.9 million a year. Those are large companies, and the figure for a smaller company is different. The mechanism, however, is the same: wrong data creates rework and wrong decisions.
The shift in thinking: close a little every day
A fast close does not come from working harder in the first week. It comes from not letting things pile up. If every inflow and outflow is recorded and checked on the day it happens, the end of the month becomes a review, not a reconstruction.
A slow monthly close is a sign that the checking is being done once a month. The goal is to do it little by little, every day, preferably without anyone noticing.
That is only feasible when data moves between systems on its own. Asking someone to check everything by hand every day just swaps one problem for another.
Step by step: leaving the side spreadsheet behind
- List where each number comes from. Revenue, costs, taxes, delinquency: for each one, which system is the source. Where there are two sources, pick one.
- Standardize the categories. A short chart of accounts, with written definitions. Too many categories lead to wrong entries.
- Integrate sales and billing. A closed sale should generate the charge with no retyping. This is where most of the differences are born.
- Automate reconciliation. What landed in the bank is compared with what was expected. People look only at what did not match.
- Create the pending-items list. An entry with no category, a charge with no payment, a payment with no charge. A visible queue, worked down throughout the month.
- Set the calendar. Cutoff day, review day, delivery day. With an owner for each.
The order matters. Integrating before standardizing moves the mess from one system to the other faster, as we explain in Map the process before you automate it.
What can be automated
Without changing systems, in most cases you can automate:
- creating the charge from the sale or the contract;
- due-date and overdue reminders;
- marking the payment as received when the bank confirms it;
- issuing the invoice after payment or delivery;
- building the monthly report from data that has already been checked.
The tools connect through integrations. If the terms API and webhook are still fuzzy, we wrote a guide for decision makers: API and webhook explained for managers.
What stays with people: classifying what is new, deciding on exceptions, negotiating with overdue customers and interpreting the result.
How to know it improved
Track three numbers, month by month:
- Days to close: from the last day of the month to the delivered result.
- Corrected entries: how many needed adjustment after being recorded.
- Pending items open on cutoff day: the size of the queue left for the last minute.
Record all three before touching anything. Without a starting number, improvement is just an impression.
An example to picture it
Imagine a training school with recurring monthly fees. Today, the front office records the enrollment in one system, finance generates the invoices in another and the checking is done in a shared spreadsheet. With the enrollment generating the charge, the bank marking payments as received and a pending-items queue in plain view, the person in finance stops checking everything and looks only at what fell outside the pattern. The close becomes a review of that queue.
Where to start
At the next close, write down how long each step took and where each number came from. That note is the map of the process. Usually one or two steps concentrate most of the hours, and that is where automation starts.
delpra™'s Operational Maturity Assessment evaluates exactly this in the data dimension: whether the month's result comes from a reliable dashboard or from a week of spreadsheets.
Written by the delpra™ teamProcesses, automation and AI so operations scale efficiently.