Purchase bills still arrive as paper, PDFs and WhatsApp photos. Typing them into Tally one by one is slow, and every retyped number is a chance for a mistake. OCR (optical character recognition) reads the text from these documents so the data can go into Tally without manual typing.
How OCR-to-Tally works
- Collect the documents: scanned bills, PDF invoices, or clear phone photos.
- OCR reads the fields: supplier name and GSTIN, invoice number and date, taxable value, tax amounts and total. For bank statements it reads date, narration, debit, credit and balance.
- Review in Excel: the data lands in a sheet where you can check it, fix anything unclear and choose ledgers.
- Import into Tally: the reviewed sheet is posted as vouchers in one go.
What you can save
| Task | Typing by hand | With OCR + review |
|---|---|---|
| 50 purchase bills | 2–3 hours | about 20–30 minutes |
| One month's bank statement (300 lines) | 3–4 hours | about 15–20 minutes |
These are typical figures for illustration. Your time depends on document quality and how many items each bill has.
How to get accurate OCR results
- Prefer original PDFs from the supplier or bank over printed-and-scanned copies.
- Scan at 200–300 DPI, in colour or greyscale, with the page straight.
- For phone photos: good light, camera directly above the page, full page visible, no shadow of your hand or phone.
- Avoid stamps, signatures and pen marks on top of the figures.
- One bill per page where possible; do not photograph two bills together.
Checks to do before posting
- GSTIN: 15 characters, the first two digits are the state code. A wrong GSTIN means the bill will not match GSTR-2B.
- Invoice number and date: these drive duplicate checks and the GST month.
- Totals: taxable value + tax should equal the bill total.
- Tax split: CGST + SGST for a supplier in your state, IGST for a supplier in another state.
- Duplicates: the same bill often arrives twice, once on email and once on WhatsApp.
Bank statements: OCR plus automatic ledgers
Bank statements are the best place to start with OCR because they are long, repetitive and arrive every month. Quick2Tally reads PDF, scanned and photographed bank statements, and its Auto Ledger feature learns from your Tally data to suggest the right ledger for each bank entry, so you are not choosing ledgers line by line.
Getting started
Pick one supplier or one bank account, run a month of documents through OCR, review, and import. Once you trust the process, add the rest. Download the free trial and try it on your own documents.
