Small-shop bookkeeping gets talked about as though it were one skill. It is two, and they are not equally hard. One is recording what happened. The other is deciding what it means for what you owe. Software is genuinely good at the first, and should stay out of the second.
The recording half, which is most of it
Nearly every problem a small shop has with its books is a gap in recording rather than a gap in expertise. A sale that was never entered. An expense paid in cash and forgotten by evening. A stock count from four months ago. None of those needs an accountant to prevent — they need capture at the moment it happens.
- Sales, with the date and how the money arrived — cash, UPI or credit.
- Purchases, with what you were actually charged rather than what you expected.
- Expenses, entered when the money leaves rather than at month end.
- What each customer owes, kept as a running balance rather than reconstructed from bills.
- Stock, counted in sections often enough that the number is believable.
Keep those five current and you have done the part that actually determines whether your books are any good. Khata records all five, including an accounting journal, expenses against cash and wallet accounts, a separation between business and personal money, and a profit and loss you can read without training.
The deciding half, which is not yours to guess
Then there is the other side of the line, and it does not get easier with better software.
- Whether you are required to register, and at what turnover.
- Which tax rate applies to a particular thing you sell.
- What counts as a business expense and what does not.
- What you must file, on what date, in your country.
- Anything you put your name to.
These vary by country, by turnover and by what you sell, and they change. Software that answered them confidently would be doing you a disservice, because the confidence would be the same whether or not the answer was still current.
Why good records make an accountant cheaper
The bill from an accountant is mostly time, and the time is mostly spent turning a carrier bag of receipts into something that can be reasoned about. If you arrive with that already done, you are buying judgement instead of data entry — which is the thing you actually wanted and the thing they are actually good at.
| What you bring | What you are paying for |
|---|---|
| A bag of receipts | Someone typing, at professional rates |
| A year of complete records | Answers to questions you could not settle |
| Records with a gap in March | Reconstruction, and a caveat on the result |
The signals that you have crossed the line
- 1
You are guessing about a rate or a deadline
A guess that is wrong compounds quietly until somebody asks. This is the cheapest possible moment to make a phone call.
- 2
Your turnover is approaching a registration threshold
The rules on either side of that line differ, and being on the wrong side of it by accident is expensive.
- 3
You are about to sign something
A filing carries your name whether or not the number came from software. Have someone look at it.
- 4
You have started avoiding your own reports
That usually means you no longer trust them, and untrusted records get abandoned rather than fixed. Get someone to check the shape of them once, and start again from a number you believe.
Do the recording yourself, daily, and let it be boring. Hand over the judgement. That division holds for almost every small shop, and it does not change when the software gets better.