Ask a practice what bank data entry costs and the honest answer is usually that nobody has worked it out. It appears on no ledger, nobody invoices for it separately, and the people doing it are already on the payroll. So it reads as free.
It is not free. It is one of the largest unbilled costs in a typical practice, and the reason it stays invisible is that it is spread thinly across every client and every month. Here is how to work out your own number, and what that number tends to imply once you have it.
The calculation, in four figures
You need four numbers, and you can get all of them in ten minutes without any time-tracking system.
- Transactions per account per month. Take three real client accounts of different sizes and count them. Do not estimate; estimates in this exercise are always low.
- Seconds per transaction. Time someone doing it properly, including reading the narration, deciding the ledger, typing the amount and checking it. Fifteen to twenty-five seconds is the usual honest range once thinking time is included.
- Accounts processed each month. Not clients. A client with three bank accounts is three accounts.
- Fully loaded hourly cost of whoever does it. Salary divided by working hours, plus the overhead the firm carries per person. Do not use the raw salary figure.
Why the number always lands higher than expected
Multiply those together and the annual figure tends to surprise people. Three reasons it beats the mental estimate.
The per-transaction time is longer than anyone thinks. People estimate typing speed. The real unit of work is read, decide, type, verify, and the deciding is the slow part. A narration like "NEFT-N12345678-ACME TRADERS PVT LTD-HDFC0000123" takes a moment to resolve into a ledger even when you know the client well.
Rework is never counted. The first pass is only part of it. Reconciliation differences, ledgers created twice, entries posted to the wrong account and the hunting that follows are all the same cost, and they land in a different month from the entry that caused them.
The expensive person does the hard part. Routine lines get delegated. Ambiguous ones get escalated, which means the highest-cost person in the firm ends up deciding individual bank transactions.
Important
Do not run this calculation using the cheapest person you could theoretically hire. Use the people actually doing the work today. A number built on a hypothetical junior tells you what you wish the cost were, which is useless for a decision.
The cost that is not measured in rupees
The financial number is only half of it, and arguably the less important half.
Bank entry concentrates into exactly the weeks when the practice has the least capacity. It does not spread evenly across the year; it piles into filing season alongside everything else carrying a statutory deadline. So the real currency is not money. It is capacity, at the precise moment capacity is scarcest.
That has knock-on effects nobody attributes back to data entry. Advisory work gets deferred. New clients get turned away, or onboarded badly. Error rates rise because the same people are working longer. And the most experienced person in the firm finishes the season having spent a meaningful share of it on transcription.
Against that, a saving expressed in rupees understates the case considerably.
What the number implies once you have it
Firms that do this calculation tend to reach one of three conclusions. All three are legitimate.
Price it, or stop absorbing it
If bank entry is a real cost and it is not in your fee, you are subsidising the clients who generate the most transactions. That is precisely backwards, since those are usually the clients with the most complex affairs and the most need of your time. Either the fee reflects volume, or the volume has to come down.
Change what arrives
A large share of the cost is caused by the input rather than the process. A scanned printout costs several times what a downloaded statement costs to handle, and clients send scans because nobody told them not to. That is a one-line fix in your document request. See scanned versus text PDFs.
Automate the mechanical part
The deciding is skilled work. The transcription is not. Software that reads the statement, classifies each line and names the counterparty removes the transcription and leaves your staff reviewing decisions, which is what they are actually being paid for.
Comparing that against software, honestly
When you do compare, compare conservatively, because the arithmetic is usually so lopsided that overstating it damages your own case.
Count only the time genuinely removed. Reviewing a classified statement still takes time, and a tool claiming otherwise is describing something you should not trust with a client's books. Count the licence at its real annual cost, including per-machine charges and whatever GST applies. And count the switching cost: a period running both ways, cleaning up the ledger master list, training people.
Even done against yourself the comparison is rarely close, which is exactly why it is worth doing that way. A number you have deliberately built to be unflattering is a number you can defend to a partner.
Pro tip
Do the calculation for one real client rather than for the whole firm. A single concrete account is far more persuasive, to yourself and to a partner, than a practice-wide estimate everyone can argue with.
Conclusion
Bank entry costs the most in the weeks you can least afford it, and its invisibility on the ledger is exactly what lets it stay that size.
Work out your own figure with the four inputs above, using the people who really do the work. Then decide deliberately: price it, change what clients send, or remove the transcription.
Greenote exists to remove the third. It reads the statement on your own PC, classifies every line, names the counterparty and posts into Tally, leaving you the review rather than the typing. See six ways to cut bank-entry time.
Start a free trial and time it against your own process.
Turn bank statements into Tally vouchers offline
Greenote reads 100+ Indian banks and posts clean vouchers straight into Tally, fully on your PC. No uploads, no cloud, no manual entry.
7-day free trial, no card required. Works with Tally Prime & ERP 9.
