How to Automate Khatabook

The udhaar that used to live in a notebook now lives on a phone

Khatabook is a free mobile app used by small businesses across India to keep a khata: the running account of what each customer owes and what is owed to each supplier. It works in Hindi, English and a dozen other languages, which is the point, because the person keeping the ledger is usually the owner and not a bookkeeper.

Udhaar is the working model of a great deal of Indian retail. The kirana that lets a family settle at the end of the month, the wholesaler supplying six shops on thirty days, the chemist, the hardware store. Before the app that was a diary with a rubber band around it, and the balance was whatever the two sides agreed it was.

What sits inside Khatabook is every credit and every payment, per customer, dated, backed up to the cloud rather than to a drawer that floods. For a shop that size it is the only record there is, and it is the one both sides trust.

The entries go in easily; the money still has to be found

Writing the khata takes seconds. Getting it settled is the rest of the month.

A customer pays by UPI and says nothing about it. The credit appears under a name that is not the customer's name, and somebody has to work out which of forty balances it belongs to before the next reminder goes out to a person who has already paid. That is the mistake that costs a customer.

A cheque was deposited on Monday, and a bounced cheque is not news that arrives. Somebody goes and looks for it.

Then the buying side, which the khata never sees at all. The distributor's rate this month against what was paid last month. Which regular has not walked in since the middle of June. Which customer with four entries already open should not be given a fifth.

Every one of those checks is a screen in somebody else's app, opened with a thumb, between customers.

Khatabook

Nobody chased for money already paid Yesterday's bank and UPI credits read and matched to the open balances, so a customer who settled on Sunday comes off the list.

A month-end outstanding list, already sorted Who owes, how much and for how long, oldest first, assembled before the shutters go up rather than after closing.

Credit history checked before more credit is given What a repeat customer has run up and how long they took last time, so the answer is a number rather than a feeling.

Khatabook already chases; it just cannot confirm

Khatabook automates the collection side better than most shops realise, and nothing below is a reason to switch any of it off.

Payment reminders go out automatically, one the day before the repayment date and one on the day itself, in the customer's own language, over WhatsApp or SMS or as a call, sent from the business account rather than from a personal number. Bulk reminders do the same for a whole list at once, turning an evening of messages into a few taps. A QR code and UPI collection put money straight into the bank account rather than a wallet somebody then has to move. Reports come out as a PDF you can hand to a customer who disputes a figure.

All of that works on entries somebody typed in. Khatabook knows the reminder went out. It does not know whether the money came, because the money lands in a bank account and a payments app it cannot see, under a reference that matches nothing. The distributor's price list, the tax portal and the shop down the road are further away still.

A payment that already landed should stop the next reminder

The bank knows things the khata does not, and it will tell anyone who signs in.

Yesterday's credits read and matched against the open balances, so the customer who paid on Sunday is off the chase list before Monday's reminder reaches them. The reference that means nothing turned into a name by amount and timing, which is what a person does by hand and what a person gets wrong on the fortieth one.

Cheques followed to cleared or bounced, so a bounce is a phone call the same afternoon rather than a discovery a fortnight later.

The rest follows the shape of the shop. A month-end list of who is outstanding and by how long, oldest at the top, ready before the shutters go up. A repeat customer's history read before more credit is extended, so the answer is a number and not a feeling. This month's distributor rate held up against the last one you were charged, and the items that went up named.

The ledger is yours; the checking does not have to be

Every morning starts the same way: the bank on one screen, the ledger on the other, and a list of credits to match between them. Nobody started a shop in order to do it.

WebRun is an agent that works a real Chrome browser, signed in as you. It opens your bank, a payments dashboard or a distributor's site, reads what each shows, matches it against what your books say is outstanding, and hands back a short list: who paid, who did not, what bounced.

It runs on your schedule in your own private environment, and any run can be watched and stopped, and the matching itself runs unattended. A reminder to a customer or a change to a balance is drafted and waits for you, because a message sent to someone who has already paid is the one mistake that loses a customer.

The workflows below are already built, and each one names what it opens.