Running a shop's udhaar book without it becoming bad debt

A customer takes ₹1,450 of goods on 3 October and pays ₹1,000 on the 11th. Write one line, "Ramesh ₹450", and you have thrown away the two facts that settle every later argument: what he took, and what he handed over. When he says three weeks later that he paid ₹1,200, you have nothing but a figure in your own handwriting.

So write two lines. On the 3rd, an entry that increases what he owes: ₹1,450, with a note naming the goods. On the 11th, a separate entry that reduces it: ₹1,000 received, cash or UPI. The ₹450 is then something the book calculates, not something you claim. Never swap the two directions. Book his ₹1,000 as goods given and he reads ₹2,450 instead of ₹450, so the book is out by twice the payment. When a balance looks wrong, halve the gap and look for a payment of that size.

Give each entry the date the thing happened, even if you write it two days later. Otherwise Saturday's closing never matches the drawer.

Slips tell you nothing until you add them up, and you add them up only when there is already a problem. A page per customer with a running balance answers the question before anyone asks it: one for Lakshmi, one for the hotel on the corner, one for the autowala who buys on udhaar.

The gain shows up in a dispute. With slips the argument is about the whole month. With a running balance you both read the same column and it narrows to one line: the 14th, ₹700, two tins of oil.

A paper page cannot tell you who wrote a line. If a son after college or a helper on Sundays also sits at the counter, they have to be inside the same ledger rather than keeping their own notes. In Khatamel that shared ledger is a group: every entry carries the name of whoever added it and the time, and the owner decides per person whether they may add, edit or delete. Let the helper add entries and keep edit and delete for yourself.

Bad debt is rarely one big sale. It is ₹400 of groceries a week for four months, two payments of ₹1,000 somewhere in the middle, and nobody doing the subtraction until the balance is ₹4,800.

Before the balance matters, write a figure beside each name. A household customer whose weekly bill runs about ₹600 can carry ₹3,000 comfortably. The hotel that buys twice a week and settles on the 5th might sit at ₹8,000. A new face gets ₹500 and one month of watching.

Three things beside each name are enough to run this well:

• The ceiling, the balance past which you collect before more goods go out.
• When their money lands: the 1st, the 5th, every Saturday evening, the week the mandi pays.
• The longest they have ever taken to clear, the honest figure rather than the one they promise.

The ceiling is a trigger for collection, not a shutter. When a ₹3,000 customer reaches ₹3,200 and wants another ₹500 of goods, refusing him in front of other people costs you the customer and leaves the ₹3,200 where it was. Say what the balance is, ask what he can give today, and hand over the goods. A payment of any size resets the date of his last payment.

₹900 that has not moved since July is a worse account than ₹4,000 taken last week. The ₹4,000 is trade; the ₹900 is a man who has quietly decided not to pay you. A balance's size tells you what is at stake; the date money last came in tells you whether you will see it.

So once a month read the page by that date, not by who owes the most. Under thirty days is ordinary udhaar. Sixty to ninety with part payments still arriving is slow and survivable, provided you have stopped adding to it. Past ninety with nothing since, and the balance has stopped behaving like credit.

Eleven such names at ₹2,000 to ₹3,000 each is ₹28,000 standing out in the lane, close to a fortnight's buying for a small kirana. Your wholesaler does not wait while you wait. Knowing it is ₹28,000 and not "a few thousand" decides whether the Diwali order comes out of what you collect or out of your own pocket.

Old accounts come back in instalments or not at all. A man owing ₹3,400 since February will not hand you ₹3,400; he will hand you ₹300 a week if you ask every week, and by the third month it is gone. Take the ₹300, write each one on the day it arrives, and keep selling him for cash so he keeps walking in. His promise of ₹500 for Tuesday takes nothing off the balance until Tuesday's ₹500 is in the drawer.

Where an account has not moved in a year and the man no longer comes, write today's date under his last line with the figure and the word closed, and stop counting it as money you have. Leave the lines on the page; take the amount out of the total you plan against. Carrying ₹6,000 of two-year-old udhaar inside your outstanding is how a book stops being worth reading.

In the ten days before Diwali a customer whose balance normally sits around ₹800 can reach ₹6,000 without one purchase you would have refused. Five hundred on Tuesday, four hundred on Wednesday, six hundred on Friday, and seven more days like them: roughly ₹500 a day for ten days, ₹5,200 on top of the ₹800 he already owed.

He is spending against a bonus not yet handed over, or a crop not yet sold. That money is real; it is why you gave the credit. But it arrives on one date and is spent inside a week, and if you are not asking in that week you are behind the jeweller, the tailor and the sweet shop. Find out which date it is for each regular before the season starts.

During those weeks, check the balance before the goods leave the counter, not at month end. It is the difference between noticing at ₹2,400 and noticing at ₹6,000. At ₹2,400 you can ask for ₹400 today and let him take what he needs for the week, and most will manage that. On the morning of Diwali, at ₹6,000, there is no figure he can produce.

Tag both sides of the season with the same word: the goods going out in October, and every payment that comes against them in November, December and January. A day-wise report filtered to that one tag then shows what the season cost you in credit and how much has come back. In Khatamel that report filters by tag over a date range, and on a paid plan you can export it as a PDF. Next year's ceiling comes from a number rather than a feeling.

The expensive mistake is waiting. The first time you mention a six-week-old ₹3,000 you are already irritated, and irritation is the one tone that turns a slow payer into a man who stops coming. A current book lets you ask flatly, because you are reading off a page: "On the 3rd you took ₹1,450, on the 11th you paid ₹1,000, so ₹450 is left. Add it to this week's order, or clear it today?"

Ask on a day he has money: the day after wages are handed out, the first week of the month for anyone on a salary. Not the day before his own rent falls due, and not the day your bazaar is shut. If you do not know when a regular's money lands, ask once and write it next to his name. Ask for a part when the whole is unrealistic, ₹500 against ₹2,300, with the date for the rest written into the note. If he has never cleared in under forty days, treat a promise for the 2nd as the 12th and ask on the 5th.

Sometimes the answer is to stop the credit. The mistake is stopping the relationship with it. Tell him the balance has to come down before more goes out, then keep selling him whatever he wants for cash. A customer who pays cash for a year and slowly clears ₹3,000 is worth more than one who owes ₹3,000 and shops elsewhere out of embarrassment.