The report lands 2 weeks after close.
By then the purchase orders it should have shaped went out a week ago, and the price changes were signed on last month's margin.
Ask where the 2 weeks went and the answer is rarely the analysis. Reading the numbers takes an afternoon. The rest of the time goes to getting the numbers into a state where they can be read.
Where the days actually go
Follow one month-end inventory report from start to finish and the time sits in 4 handoffs:
- The exports. Stock from the inventory system, balances from the accounting system or ERP, sales, purchases, sometimes a production file. Each comes out in its own layout, on its own day, from a different person.
- The tagging. Every item has to go in the right bucket: raw material or finished goods, product line, active or discontinued. In most businesses the rules for that live in one person's head, and it takes about 3 days every month.
- The matching. The same product has one name in the warehouse and another in the books. Someone lines them up by hand every month, and the list of exceptions is never quite the same twice.
- The chasing. The two systems do not agree. Someone finds the difference, explains it, adjusts it, and starts the report again.
None of this is analysis. It is the same work done again every month by the one person who knows how, and it is why the report waits when that person is busy or away.
Why a better spreadsheet does not fix it
The usual answer is a better template, a bigger macro or a new dashboard tool. Each one speeds up the last step, the building, and leaves the 4 handoffs where they were. The dashboard looks faster and still waits for the cleanup.
The deeper problem is that the rules were never written down: which date decides the period, which items count as finished goods, how a name in one system maps to a name in the other. When the rules live in people, two people produce two reports.
An automotive-related company found exactly that. Finance and operations worked from the same source data and reached different numbers, because each team had built its own cost rules over the years and neither knew the other had moved. Once one set of rules was written and used by both, the reporting discrepancies went away and the accounting adjustments fell.
"Both teams were right by their own logic. That was exactly the problem."
What changes when the rules are written once
An AI workflow does the 4 handoffs the same way every month. It pulls the exports, cleans them, classifies every item, separates raw material from finished goods, matches names across both systems, runs the checks and builds the report. The rules are written once, in plain language the team can read and change. When something does not tie, the workflow stops and names it instead of passing a wrong number forward.
The judgment stays with people. The repeated work does not.
The CFO gets the report days after close and not weeks. It opens on what needs attention this month, and every figure traces to the transactions behind it. The 3 days of tagging rows go away.
A test for your next close
At the next month-end, write down 2 dates: the day the last export lands, and the day the report reaches the CFO. Then ask the person who builds it to list everything done in between.
Mark each step: judgment, or the same as last month? If more than half the list is the same as last month, the report is waiting on handoffs, not on analysis.
See it on one month
Northvale Furniture makes chairs, bookcases and cabinets for retailers and contract buyers: 8 product lines, one warehouse, month-end done in-house. Its CFO dashboard shows the month on one screen: $264,590 sitting still for 180 days inside a month that looked fine on the P&L, the exceptions ranked by what they are worth, every figure traced.
Is this your number?
If the report takes days of manual work before anyone can read it, this is the work: one clean report from systems that do not talk.