A pos cash drawer printer setup is usually judged at installation. Choosing the pair is covered in choosing the pair. This page judges it at closing time.
Key Takeaways
- A close is only as good as the setup behind it.
- Count the float onto a written figure, twice daily.
- Cash and digital must be shown separately.
- Tie every drawer opening to a document.
- Print the summary; do not read it off a screen.
- Four of five differences are not theft.
What a pos cash drawer printer setup has to prove each night
Closing the drawer is treated as a chore, and it is the only daily check a shop performs on itself.
The purpose is narrow. Counted cash should equal the opening float plus recorded cash sales, less any cash paid out.
If those two figures agree, the day is confirmed. If they disagree, the shop has learned something while the day is still fresh enough to investigate.
A close that cannot produce those two figures is not a check at all, and most shops discover this only after a difference appears.
The ability to produce them is decided by the setup rather than by the diligence of whoever is counting.
Rule 46 of the CGST Rules lists the invoice particulars, which is what makes the day’s sales a set of documents rather than a running total on a screen.
Decision one: the drawer follows a document
The first setup decision is the one that makes counting meaningful at all.
Drawer kick documentation describes how the printer releases the drawer, which means a drawer opening can be tied to a printed sale.
Configured that way, every opening has a document behind it and the number of openings should match the number of cash transactions.
Left unconfigured, the drawer opens whenever somebody presses a button, and the count at closing has nothing to be compared against.
This single decision is what converts a cash box into a controlled drawer, and it costs nothing beyond asking for it.
The reasoning behind the connection is in why they are connected.
Decision two: the no sale needs a name
Every system permits a drawer opening with no transaction, and most shops never restrict it.
Genuine reasons exist. Giving change for a note, correcting a float, retrieving something left in the drawer.
The problem is not that it exists but that it is anonymous, which makes a whole category of openings uncountable.
Requiring a login turns each one into a line with a name and a time against it, which is a record rather than an accusation.
Read the no sale list weekly. In most shops it should be close to empty, and a list that grows steadily is the earliest useful warning.
Keep the function rather than removing it, because a counter that cannot open its drawer legitimately will find another way.
- Keep the no sale function.
- Require a login for every one.
- Read the list weekly, not monthly.
- Treat steady growth as the signal.
Decision three: cash and digital shown apart
The commonest reason a close fails is that it compares a cash count against a total that was never all cash.
UPI settles account to account, so digital takings never pass through the drawer at any point.
A day summary that reports one figure forces whoever is counting to subtract the digital share by hand, which introduces exactly the error the close exists to catch.
Split payments make it worse, because a single bill settled half in cash and half digitally belongs partly in each column.
Confirm at installation that the summary separates the two, and confirm that a split payment appears correctly in both.
This is the one report an absent owner will actually read, so it is worth insisting on its shape.
Decision four: a float that does not drift
The float is the baseline, and a baseline nobody writes down is not a baseline.
Fix the amount and keep it the same every day, because a float that varies makes every difference impossible to place.
Count it at opening onto a written figure, and count it again at closing onto the same sheet.
A float topped up from the day’s takings without a note is the single commonest cause of a shortfall that looks sinister and is not.
RBI guidance covers the handling and exchange of soiled and damaged notes, which a counter accumulates and which should leave the float rather than sit in it.
The daily routine around the pair is set out in running the pair day to day.
Decision five: the summary prints
A figure read off a screen is a figure nobody can check tomorrow.
Printer paper width decides legibility, and 80mm keeps a closing summary readable with cash and digital in separate columns.
Print it, write the counted cash on it by hand, and file the two together.
That single habit turns the close into a record. A month of filed summaries shows whether differences are random or patterned.
Random differences across cashiers and days point at procedure. Patterned ones point somewhere more specific.
Neither conclusion is available to a shop that reads a number, nods and closes the screen.
| On the printed summary | Written on by hand | What it lets you see |
|---|---|---|
| Recorded cash sales | Counted cash | The difference, plainly |
| Digital takings | Nothing | The bank comparison |
| Opening float | Closing float | Whether the baseline moved |
| No sale count | A reason for each | Unattached openings |
| Who was on shift | Their signature | Who to ask tomorrow |
Decision six: one drawer per cashier
The last decision applies only where more than one person serves on a shift, and it is absolute where it applies.
A shared drawer makes every login record unattributable, because two people had their hands in the same cash.
This matters most when something goes wrong, since a shortfall on a shared drawer cannot be placed and therefore cannot be discussed usefully.
Separate drawers protect the honest cashier far more than they catch the dishonest one, which is how the change should be explained.
Where a second drawer is not affordable, the alternative is a single named person on the drawer per shift, handed over with a count.
A handover count takes two minutes and converts an ambiguous day into two accountable halves.
- One drawer per cashier where you can.
- Otherwise one named person per shift.
- Count at every handover, in writing.
- Both parties sign the handover figure.
- File it with the day’s summary.
The weekly read, which catches what a night cannot
A single night tells you very little, and a month of filed summaries tells you almost everything.
Set aside fifteen minutes once a week and lay the summaries out together rather than reading them one at a time.
Look first at whether the differences cluster on particular days, because a Saturday pattern is a queue problem rather than a person problem.
Look next at whether they cluster on particular people, which is worth noticing quietly and worth checking against who was simply on the busiest shifts.
Look at the no sale list across the week, since a single day’s list is too short to show anything and a week’s is not.
Finally compare the digital column against what actually arrived in the bank, because that is the half of the day a drawer count never touches.
| Weekly pattern | What it usually means | What to change |
|---|---|---|
| Differences on busy days | Change given under pressure | A second position |
| Differences on one person | Check their shifts first | Then a conversation |
| No sale list growing | The control has lapsed | Re-explain, do not re-restrict |
| Digital does not match bank | A tender recorded wrongly | Check split payments |
| Nothing clusters at all | Ordinary noise | Nothing |
The last row matters as much as the others. A counter with small random differences and no pattern is a counter working normally.
Chasing that noise costs goodwill and finds nothing, which is why the weekly read is more useful than a nightly interrogation.
- Read a week together, never one night alone.
- Check busy days before checking people.
- Compare digital against the bank weekly.
- Accept random noise without chasing it.
What the counter should be told
A closing routine introduced badly reads as suspicion, and a counter that feels suspected becomes less careful rather than more.
Explain what the close is for before changing anything, in the order the staff will experience it.
Say that the dull causes are checked first, and name them: the float, the change given, a sale not rung up, a tender recorded wrongly.
Say that a difference is a question about the day rather than about the person, and then behave that way on the first difference.
Point out what the routine protects. A cashier on a separate drawer cannot be blamed for somebody else’s shortfall, and that is a real benefit to them.
Say that the summary is filed and reviewed weekly rather than nightly, so a single odd night will not produce a conversation.
A counter told all of that in advance handles the first difference calmly, which is the whole point of saying it.
- Explain the routine before changing anything.
- Name the dull causes out loud.
- Check them first on the first real difference.
- Point out that separate drawers protect them.
- Review weekly, not nightly.
Why a pos cash drawer printer setup decides this and not diligence
It is worth being explicit about why six installation decisions matter more than how carefully anybody counts at the end of a long trading day.
A careful person working with a mixed total and a drifting float cannot produce a meaningful figure, however much attention they bring to the task.
A careless person working with separated tenders, a fixed float and attached openings will still produce a figure that can be checked by somebody else.
The setup determines what is knowable; the person determines only how quickly it is written down. That asymmetry is why the six decisions repay the afternoon they cost.
It also explains why shops that train harder on counting rarely improve, while shops that change one setting frequently do.
Where a counter has never reconciled, the honest first question is which of the six decisions was never made, rather than who was on the drawer.
Audit the pos cash drawer printer setup against the six before anybody is asked to count more carefully, because five of the six are settings rather than purchases.
- Setup decides what can be known.
- Diligence decides only the speed.
- Training harder rarely fixes a mixed total.
- Ask which decision was skipped, not who counted.
The closing routine itself, in order
With the six decisions in place, the routine takes about ten minutes and never varies.
Print the day’s summary first, before touching the cash, so the recorded figure is fixed before anybody knows the counted one.
Count the drawer without looking at the summary, because a counter who knows the target will unconsciously count towards it.
Write the counted figure on the printed summary, then compare.
Remove the day’s takings and recount the float to its fixed amount, writing that figure down as well.
Investigate any difference the same night, while the staff who worked the day are still present and the hours are still remembered.
File the summary. A difference explained and filed is a closed matter; one absorbed silently becomes a pattern nobody can describe.
- Print the summary before touching cash.
- Count the drawer without reading it.
- Write the counted figure on the summary.
- Recount the float to its fixed amount.
- Investigate differences the same night.
- File the summary with the count on it.
Reading a difference without accusing anybody
Most differences have dull explanations, and starting with the dull ones keeps the counter working well.
A shortfall equal to a round figure is almost always the float, miscounted at opening or topped up without a note.
A shortfall in odd amounts across a busy day is change given wrong, which is a function of the queue rather than the cashier.
A surplus is more interesting than a shortfall, because it usually means a sale was not rung up at all.
A shortfall exactly equal to one bill is frequently a digital payment recorded as cash, which the tender split would have prevented.
Only a repeated shortfall, with the same person, across weeks, points anywhere uncomfortable, and even then the override list should be read alongside it.
Say this order out loud to staff. A counter that knows the dull explanations are checked first behaves very differently from one that expects suspicion.
Fixing a pos cash drawer printer setup that cannot close
An existing counter that never reconciles can usually be corrected in one session.
Start with the summary, because a report that mixes cash and digital makes everything else pointless.
Then tie the drawer to the printer, which is a setting rather than a purchase on most counters.
Then restrict the no sale to a login, and accept that the first week will produce a surprisingly long list.
Then fix the float at a round amount and start the written count sheet, which is a notebook rather than a system.
Do not attempt all four during a busy week. One change per week, in that order, with the counter told why each one is happening.
| Week | Change | What it fixes |
|---|---|---|
| One | Separate cash and digital | The comparison itself |
| Two | Tie the drawer to the printer | Uncountable openings |
| Three | No sale requires a login | Anonymous openings |
| Four | Fixed float, written count | A drifting baseline |
Wrong fit: closing assumptions that hide problems
Five assumptions recur. Each is a wrong fit, with the honest version beside it.
| Assumption | Why it is a wrong fit | Honest version |
|---|---|---|
| The total is the total | Wrong fit: digital never entered the drawer | Split the two columns |
| The float is roughly right | Wrong fit: a drifting baseline hides everything | Fixed, and counted in writing |
| A small difference is fine | Wrong fit: patterns are built from small ones | File every night |
| No sale is harmless | Wrong fit: it makes openings uncountable | Tie it to a login |
| A difference means theft | Wrong fit: four causes come first | Check the dull ones first |
The installation errors behind most of these are in the setup errors behind a bad close.
Judged at closing time, a pos cash drawer printer setup is good or bad long before anybody counts a note.
Terms a close uses
Four expressions recur, and settling them makes the nightly conversation shorter.
A float denotes the fixed cash a drawer starts with, which should be identical every morning.
A tender denotes a payment method on a bill, which is why a split payment is one sale with two tenders.
A no sale denotes a drawer opening with no transaction behind it.
A variance denotes the gap between counted cash and recorded cash, which is a figure to explain rather than a verdict.
- Float: fixed opening cash, counted twice.
- Tender: a payment method on a bill.
- No sale: an opening with no transaction.
- Variance: the gap, to be explained.
Conclusion
Tie the drawer to a document, name every no sale, split cash from digital, fix the float, print the summary and give each cashier a drawer. Then the ten minute routine actually proves something. Keeping the pair healthy is covered in keeping the pair working.
Sources and method
Device information comes from maker docs. Star Micronics support pages describe drawer kick behaviour, and Epson India publishes paper width. The rules are primary. NPCI documents how UPI settles, Rule 46 of the CGST Rules sets the invoice particulars, and the RBI publishes guidance on handling banknotes.
What we could not verify: no figures are given for typical till variance, because none is published and it depends entirely on volume and procedure. The ordering of causes comes from counter practice.
This is general guidance, not accounting advice. Confirm cash handling and record keeping requirements with your accountant.
Written by Aman raj for Clonet Technologies, a manufacturer of printers, weighing scales and point-of-sale equipment.
Frequently asked questions
Why does my till never balance?
Usually the float, the tender split or unattached drawer openings. Four of the five common causes are setup and procedure rather than anybody taking money.
Should digital payments be counted in the drawer?
No. UPI settles account to account, so digital takings never enter the drawer. The day’s summary has to show cash and digital separately.
What is a no sale, and why does it matter?
A drawer opening not attached to a transaction. If anybody can trigger one, drawer openings stop being countable and the close loses its meaning.
How should the opening float be handled?
Fixed amount, counted onto a written figure at open and at close. A float that drifts makes every subsequent difference impossible to place.
Does each cashier need their own drawer?
Where more than one person serves on a shift, yes. A shared drawer makes every login record unattributable in practice.