Nine in the evening, shutters half down, and the owner is counting notes into piles on the glass counter while a staff member waits by the door wanting to go home. The count comes to Rs 1,400 less than it feels like it should. Nobody knows whether that is a mistake, a memory, or a person.
That silence is the most expensive minute in a small shop, and almost every shop in Nepal has stood in it.
The problem is not theft, it is not knowing
Owners tend to jump straight to theft, and then feel guilty for jumping there, and then say nothing. That is the worst of all outcomes: the money is gone, the staff member can feel they are suspected, and no information was gained.
The real problem is that the shop has no figure to count against. Cash came in all day from a dozen small sales, some of which were rung up, some of which were "just add it later", some of which were a customer paying off part of an old khata balance. At nine o'clock there is a pile of notes and a vague sense of how the day went. A vague sense cannot be reconciled against anything.
So you are not really looking for a thief. You are looking for a number that should have existed at five in the afternoon and did not.
What a close actually does
Closing the day properly is one small ceremony: the till tells you what it thinks it took, you count what is actually there, and the difference gets written down.
That is it. The magic is not in the counting, it is in the difference being a recorded number rather than a feeling. Rs 1,400 short on Saturday is an event. Rs 60 short most days and Rs 1,400 short on one Saturday is a story, and the story is what you need.
In Pasal the counter has a Close day step that does this arithmetic for you, and it does it against the sales the till actually rang, including the ones taken while the internet was down. A sale that could not reach the server completes for the customer anyway and syncs when the connection comes back, so a dead wifi hour does not quietly fall out of the day's figure. That matters here more than it would elsewhere.
Small and often beats big and never
The reason most shops do not close the day is that it feels like accounting, and accounting is something you do later.
But a close is not accounting. It is two minutes, and its whole value comes from being done every single day. A month of daily closes tells you which shift runs short. A single close at month end tells you nothing at all, because a month is far too long a window to locate anything inside.
If you are starting from nothing, do not begin by reconstructing the last three months. Begin tonight. Tomorrow you will have one day of truth, and in a fortnight you will have a pattern, and a pattern is the thing that lets you have a calm conversation instead of an accusation.
The moment you hire one person, names matter
While you are alone behind the counter, a shortfall is your own problem and you will find it eventually.
The day you hire somebody, an unattributed sale becomes an unanswerable question. Not because your staff are dishonest, but because "who rang this up" is information the shop needs in order to be fair to them. A staff member who is trusted and can prove it is in a much stronger position than one who is trusted vaguely.
This is why the till in Pasal asks who is ringing, with a PIN checked on the server rather than in the browser, and stamps the name onto the sale. It is not surveillance. It is the same reason a khata entry has a date on it. When a customer disputes something in three weeks, or when a drawer is short, the record answers instead of two people remembering differently.
And it cuts the other way, which is the part worth saying out loud to the person you hire. A named sale protects the seller. Without names, every shortfall lands on whoever happened to be there.
Do not let the drawer be the only truth
Two habits sit alongside the close and make it worth much more.
Ring everything, including the small stuff. A shop where two hundred rupee sales are sometimes not rung up will never reconcile, and the person who skips them is usually doing it to be quick rather than to be dishonest. It is worth saying explicitly that speed is not the goal at the counter, completeness is.
Keep khata payments out of the day's sales in your head. Somebody paying off Rs 3,000 of old credit is cash in the drawer that is not a sale today, and a shop that mixes those two will conclude its takings are wonderful and its stock is haunted.
The end of the argument
The point of all this is not control. It is that nobody should have to stand at a counter at nine at night with a pile of notes and a suspicion.
A day that closes against a real figure turns that minute into a number written down, and a number can be looked at again tomorrow when everybody is less tired. That is a better shop to own and a considerably better shop to work in.
The staff member by the door wants to go home. Give them a close that takes two minutes and a record that says the day was right, and they will.
The two minute version at the counter is closing the day.