Daily Takings & Till Reconciliation

What this tool does

CM8-113 is a daily takings and till reconciliation sheet for a shop, cafe, bar, salon or any other counter where cash still crosses it. You enter what the till says you sold, what you counted in the drawer and what went out of it; the tool works out how much cash should have been there, compares it with how much actually was, and shows you which days are far enough out to be worth explaining.

Everything runs inside this one file. There is no account, no upload and no network request of any kind, so your takings, your operators' names and your banking references never leave the computer you are using.

One record is one till, for one shift, on one day. If you have three tills and count each of them at close, that is three records for that date — and three records is what the counts in this tool are counting. If a drawer is counted twice a day at changeover, record each count separately with its own float.

What to record each day

• Opening float — the cash put in the drawer before trading started.
• Cash counted at close — everything in the drawer when you finished, including the float. This is the figure people most often get wrong: if you enter the cash after removing the float, every day will look short by exactly the float.
• Card takings — the terminal's own total for the same period, not the card figure the till reports. Where the two disagree, the terminal is usually right about what was taken and the till is right about what was rung up; that gap is worth a separate look.
• Other tender takings — vouchers, gift cards, account sales, mobile and app payments. Anything that is neither cash nor card belongs here, or the cash figure will be wrong.
• Till reading — the Z-read or system total for all tenders together. Not the cash element. This is the single most important entry: every derived figure hangs off it.
• Cash paid out of the till — petty cash taken from the drawer during trading.
• Banked amount and reference — what physically went to the bank, and the slip or bag number that proves it.

How the reconciliation works

Three steps, all of them arithmetic you could do on paper. The tool shows each one so anybody can check it.

The till reading is everything you sold. Take off the tenders that were not cash and what is left is what should have come across the counter as cash.

Add back the float you started with, take off anything you paid out of the drawer, and that is what should be sitting there when you count.

A positive variance is an over: more cash in the drawer than the till accounts for. A negative variance is a short. Both are errors. An over is not good news — it usually means a sale was never rung up, which is the same control failure as a short, only pointing the other way.

The percentage puts a variance in proportion. The same short means something very different on a quiet Tuesday than on your busiest day of the year.

Refunds, pay-outs and floats

Refunds are recorded for context and are not in the formula above. The tool assumes your till reading is already net of refunds, which is how most systems report. If yours reports gross sales with refunds shown separately, subtract them yourself before entering the till reading — otherwise every day with a cash refund will show as short by the amount refunded.

Pay-outs are the other frequent cause of a false short. Money taken out of the drawer for a delivery, a taxi or a window cleaner has to be recorded here, with the receipt kept in the drawer until the count. A pay-out that nobody wrote down is indistinguishable from missing cash.

Float top-ups are the mirror image and produce false overs. If somebody adds change from the safe mid-shift without recording it, the drawer will count high by exactly that amount. Either record the top-up as part of the opening float and note it, or run a float sheet that the count is checked against.

Setting a sensible tolerance

No till balances to zero every day, and chasing a small variance costs more than it recovers. The tool therefore flags a count only when it fails both tests you set:

The cash tolerance stops trivial amounts being escalated whatever the day's size. The percentage lifts the bar on a busy day, where a slightly larger variance is proportionately smaller and much more likely to be ordinary handling error. Set either to zero to rely on the other alone; set both to zero and every non-zero variance is flagged.

Reasonable starting points are a cash tolerance of about a fifth of your average transaction value and a percentage of a half to one per cent of takings. Then watch what actually gets flagged for a fortnight and adjust. A tolerance that flags every day teaches people to ignore the flag; a tolerance that flags nothing is not a control at all.

Why the net figure lies

The net variance is the sum of every over and every short, so they cancel. A month of ten shorts of 20 and ten overs of 20 nets to zero and looks perfect, while 400 has moved in and out of your drawers unexplained. That is why the Net variance tile always shows the total absolute variance underneath it, and why the operator summary shows both columns side by side.

Read them together. A net near zero with a small absolute total means your counting is genuinely tight. A net near zero with a large absolute total means your counting is loose in both directions, which is the condition in which a real loss can hide unnoticed.

Reading a pattern — and what it never proves

A single variance is almost always a counting or keying error. Somebody transposed two digits, gave the wrong change once, rang a sale as cash that was taken on card, or miscounted a bag of coin. Investigating one day in isolation usually wastes an hour and finds nothing.

What is worth attention is a pattern: the same till short most days, or the same operator short across several tills, or shorts that cluster on one shift. The per-till chart and the operator summary exist for exactly that, and the note field is there to record what you found.

The second-person check field records whether somebody other than the operator counted or verified the drawer. That, not this tool, is the control. A variance found by the person who created it is a much weaker piece of evidence than one found by somebody else, and separating the two roles removes most of the awkward conversations before they start.

The usual causes of a variance

Before escalating anything, work down this list. It resolves most of them:

• The float was counted out of the closing cash, or topped up without a record.
• A pay-out was made and the receipt was lost or never entered.
• A card sale was keyed as cash, or the reverse — this produces a large variance with a perfectly correct total, so check the tender split before you check the count.
• The card terminal total covers a different cut-off time from the till reading.
• A voucher or gift card was rung through as cash, or not recorded at all.
• Change was given from the wrong denomination, most often a note mistaken for another.
• A refund was given in cash but processed as a card refund, or not put through the till at all.
• Two tills share a drawer, or trade was moved between tills mid-shift.
• The count itself was rushed: bagged coin taken at its printed value without weighing or counting.

Banking and the paper trail

Record what actually went to the bank and the slip or bag number that proves it. Banking is a separate question from reconciliation — you may bank two days together, or hold cash back for change — so the tool does not force the banked amount to agree with the cash counted. What it can do is insist you never bank without a reference: turn on the requirement in Settings and a record with an amount banked and no reference will be refused.

If you bank several days or several tills in one deposit, put the whole amount against one record and say so in the note, so the reference can still be traced from either end.

Reading the charts and tables

The tender donut shows how customers paid, using the till's own figures. The cash slice is cash sales, not the cash you counted — the difference between those two is the whole point of the exercise. Watching this split move over a year tells you when to renegotiate card charges or reduce the banking round.

The takings bar chart switches automatically: sixteen trading days or fewer are drawn day by day, and beyond that it groups by month so the labels stay legible.

The variance trend plots every till on a date added together, so a day with an over on one till and a short on another can appear flat. The dashed line is your cash tolerance; the same distance below the axis zero is the other side of it.

The per-till bars show the size of the net variance, so a short and an over of equal size draw the same length of bar — the sign and the absolute total are in the text alongside. The counts outside tolerance table lists every flagged day with the expected figure, the counted figure and the variance, and tells you whether a note was written. The operator summary puts net and absolute variance side by side per person.

Printing and sharing

Print Report produces a report from whatever the current filter shows: the header, the headline figures, all four charts, both summary tables, the full register and your closing notes. Print to PDF to circulate it.

The scope line under the title states the filter in force, and the currency the figures are in. Clear the filters before issuing anything described as a full period. Remember that the report names operators against variances — think about who receives it, and consider filtering to a single till or using initials.

Saving your work

Counts, settings and the report header are written to this browser's local storage as you type, and the toolbar shows the time of the last save. That storage belongs to one browser on one computer: another browser, a private window, a second machine or a clean-up tool that clears site data will not have it.

Treat Export .json as the real save — one file containing everything, which Import .json restores anywhere. Export CSV gives you the register for spreadsheet or bookkeeping work, including the columns hidden from the on-screen table, and covers every filtered record rather than only those drawn. Reset asks twice, then erases everything this tool has stored. There is no undo.

How long you must keep records of daily takings, and in what form, differs by country and by the kind of business you run. Establish what applies to you and archive the exports accordingly — this tool keeps nothing for you.

Accuracy & disclaimer

This tool calculates from what you enter and nothing else. It cannot tell whether the till reading is right, whether the drawer was counted properly, whether the card terminal covers the same period, or whether a pay-out was recorded. Every variance it shows depends on all four.

It is an internal record-keeping and calculation aid. It is not an accounting system, not a stock or margin calculation, not a tax record in itself, and not evidence of anything about any person. Variances are differences between figures; establishing why one exists is a matter for your own investigation, carried out fairly and under whatever rules apply where you are.