Three systems, three different revenue figures.

Ask a multi-site operator what they took last month and you will often get a pause. Not because they do not know, but because they know there are three answers and are deciding which one to give you.

The POS says one number. The stock system implies another. The ledger shows a third. None of them is lying; they are measuring different things at different moments. But if nobody owns the reconciliation between them, month-end becomes an argument about which figure to believe rather than a decision about what to do.

Why the three never match

The gaps are almost always the same handful of things, and once you know them the differences stop being mysterious.

• Timing. The POS closes at 2am on the trading day. The bank settles card payments a day or three later, sometimes netted, sometimes gross. A month boundary falling mid-week guarantees a difference.
• Deductions you never see. Card fees, delivery-platform commission and refunds are frequently netted off before the money reaches you. Gross sales in the POS, net cash in the bank.
• Deferred income. Vouchers sold, deposits taken and events paid for in advance are cash today but revenue later. The POS often records them as a sale on the day they are taken.
• Staff meals, comps and wastage. These move stock without producing revenue. If the stock system assumes everything that left the shelf was sold, its implied revenue will always overstate.
• Service charge and tips. Depending on configuration, these may sit inside the POS sales figure while belonging nowhere near your revenue line.

Each has a correct treatment. The problem is not complexity — it is that nobody has been made responsible for applying them consistently, every period, in the same way.

If three systems disagree, the problem is ownership, not software. Buying a fourth system does not fix it.

The cost of not reconciling

The obvious cost is time — days of it, every month, spent by someone expensive tracing differences that recur identically next month.

The larger cost is that nobody trusts the numbers, so nobody uses them. When a GP figure is met with "that doesn't look right", the conversation stops being about margin and becomes about data. Meanwhile the actual problem the figure was pointing at goes unaddressed.

There is a third cost that surfaces later. Unreconciled revenue makes your VAT position an estimate rather than a fact, and it makes prior-period adjustments at year end much more likely — which is precisely when you least want surprises.

What fixing it looks like

Unglamorous, and mostly about consistency rather than cleverness.

1. Name one source of truth. The ledger. Every other system reconciles into it. This sounds obvious and is the step most often skipped.
2. Post daily, not monthly. A day's variance is traceable in minutes because somebody remembers the shift. A month's variance is archaeology.
3. Reconcile each payment channel separately. Cash, each card merchant, and each delivery platform. Lumping them together hides exactly the differences you need to see.
4. Treat deferred income properly at the point of sale. Vouchers and deposits go to a liability, not revenue, and are released when redeemed.
5. Agree the stock treatment for non-sales. Staff meals, comps and wastage recorded consistently, so the stock system's implied revenue is comparable to the ledger's actual revenue.

What you get back

When this is working, month-end stops being an investigation. The variance between POS and ledger sits under a fraction of a percent, and when it moves, that movement is itself information — a till discrepancy, a missed banking, a platform that changed its fee structure without telling you.

More usefully, weekly reporting becomes possible. You cannot produce a credible flash report on Tuesday if the underlying numbers are only reconciled once a month, six weeks later. Daily discipline is what makes weekly insight possible, and weekly insight is the thing that actually changes decisions.

If your systems currently disagree and nobody owns closing the gap, that is the piece worth fixing first. Almost everything else in your reporting depends on it.