“Why doesn’t sales revenue match the ledger?”
I hear this question in almost every finance team. Sales reports one figure. Finance reports another. Somebody exports both to Excel, filters, compares and asks around. That usually costs a day. In some companies it costs a day every month. And at the end, only one person knows how the answer was reached.
The data to answer it sits in Business Central. So I asked Business Central instead.
Where the gap comes from
Sales looks at revenue from the item ledger entries: what was invoiced, per item, per customer. Finance looks at the revenue accounts in the general ledger. Most months the two agree. When they don’t, the usual suspects are:
- manual journal entries on revenue accounts
- credit memos and corrections
- payment discounts and allowances
- prepayments
- postings from other sources that never touch an item
None of this is exotic. The problem is that nobody checks it systematically. The difference is found when someone notices it, often late, often under time pressure before a board meeting.
What I built
A small AL extension in Business Central that does the reconciliation on request. For the last 24 months it compares two figures per month:
- revenue according to the item ledger entries (the subledger)
- revenue on the revenue G/L accounts, split by origin: from sales documents, from manual general journal entries, from other sources
A second API returns the individual postings on revenue accounts that do not come from sales documents. These are exactly the candidates for a difference.
Which accounts count as revenue accounts? The extension reads them from the General Posting Setup. There is no extra list to maintain. If someone adds a new revenue account to the posting setup, the reconciliation picks it up.
How it works
The principle is the same throughout my “Hack Business Central” series: AL calculates, the AI explains.
- The AL extension calculates the monthly comparison and the list of non-document postings. Deterministic, testable, the same result every time.
- Business Central’s MCP server exposes these APIs as tools an AI assistant may call. The configuration is read-only.
- The AI assistant picks the right tool, reads the result and writes the answer in the language of whoever asked.
The language model does no sums. It receives finished figures and is asked to explain them. Every sentence in its answer traces back to an API and a rule.
The result in the demo
The demo runs on Microsoft’s CRONUS demo company with an invented business story. There is no client data.
I asked: “Why does August revenue differ from the sales report?”
In August 2026 revenue in the general ledger is €24,700, in the subledger €15,200. The difference of €9,500 comes entirely from a manual general journal entry on 31 August to revenue account 6110 (“Manual correction project revenue Q3”, document KRBJNL-0001). All other months match. Check whether this correction has supporting documentation, and why the account allows direct posting.
Three things in this answer matter to a finance team:
- The gap is located. One month, one account, one posting.
- The rest is confirmed. All other months match, so nobody needs to search there.
- The cause is named. Revenue account 6110 has
Direct Postingenabled. That is why a manual entry could land there at all.
That last sentence is the real win. Not just where the gap is, but why it could happen. Finding the €9,500 fixes August. Questioning direct posting on revenue accounts prevents the next September.
What it doesn’t do
The limits, honestly:
- It doesn’t decide. Whether the correction was right, and whether direct posting should be switched off, is a decision for the controller. The assistant cannot post, change or delete anything.
- It is only as good as the posting setup. If the General Posting Setup has gaps, so does the list of revenue accounts. That is worth knowing, and often a finding in itself.
- It explains the source, not the intent. The extension sees that a manual entry was made, by whom and with which document number. It cannot know whether the supporting document exists on someone’s desk.
- The language model can be wrong. That is why it never calculates. If the wording is off, the figures underneath are still the ones Business Central computed.
- Business Central online only. The MCP server does not exist on-premises. There you need a different route to the same reconciliation.
Try it
You can see the reconciliation and the AI commentary in the interactive demo: Open the Management Cockpit for Business Central. Demo data only.
The architecture, the three use cases and the full list of limits are in the white paper: Ask Business Central: three AI use cases that work today.
Is your team facing a similar problem?
If someone in your company re-does this reconciliation in Excel every month, that is a good place to start. Discuss your challenge. I usually reply within two working days.