CSV to QBO Converter for Restaurants: Import POS, Card Batch and Delivery App CSVs into QuickBooks

Five rails, five clocks, and not one of them deposits the gross amount. That is the whole reason the Z report and the bank feed disagree, and it is why the fix is not a tighter sync between the POS and QuickBooks. The two systems are measuring different things at different moments. A daily sales journal entry records what the restaurant sold. A clearing account records what the processors and marketplaces owe you. The deposits then clear against that account as they arrive, and the gap between them is just float, not an error.

The most common wrong turn is to skip the sales entry and simply categorize each deposit as income. It feels efficient and it destroys the P&L. Booking a DoorDash payout as revenue records the net figure as sales, which understates both revenue and expense by the commission amount. Booking a card batch as revenue buries the processing fee. Booking the full deposit as revenue does the opposite, counting sales tax and tips you are holding for somebody else as money you earned. Either way the top line is fiction, and every percentage you calculate from it inherits the error.

Record one summary sales entry per day that mirrors your POS daily sales summary line for line, in the same order, then memorize it as a recurring template. Debit each payment rail to its own clearing account or to Undeposited Funds for cash, credit each revenue category, and credit sales tax payable and tips payable as liabilities. The entry should balance to the POS gross, not to any deposit.

The discipline that makes this work is keeping one clearing account per rail rather than one for everything. Card batches, Amex, and each delivery marketplace get their own, because each settles on its own schedule and nets out its own deduction. When a deposit arrives, it clears against a single account and the match is unambiguous. A healthy clearing balance sits at roughly one to two days of that rail's net sales. A balance that keeps climbing means deposits are being matched somewhere else or missed entirely; a balance that goes negative means daily sales entries are being skipped.

You are not entering tickets. Nobody should be keying individual checks into QuickBooks. One entry a day, or one per week for a small operation, is the right level of detail for a general ledger. The per ticket record lives in the POS and belongs there.

Record the gross order value as sales and the platform's commission as a separate expense, then clear the net payout against the delivery clearing account. The commission is a cost of doing business, not a discount on the sale. DoorDash publishes three merchant plans at 15 percent for Basic, 25 percent for Plus and 30 percent for Premier on delivery orders, with 6 percent on pickup for Basic, so the amount involved is not a rounding error.

Consider a $1,000 week of DoorDash orders on the Plus plan. The deposit is $750. If you book $750 of revenue, you have hidden $250 of marketing and delivery cost and told yourself you sold $750. Your food cost percentage is now calculated against the wrong denominator and looks worse than it is, while your operating expenses look better than they are. On the sales tax return the problem is sharper still, because most states want tax on the gross sale to the customer, not on what the platform forwarded you. The platform's CSV report is the document that lets you split gross from commission, which is exactly why the CSV matters more here than the deposit does.

No. Tips belong in a liability account, usually called Tips Payable, because the money is owed to staff and was never the restaurant's revenue. Credit Tips Payable when the tip is added on the card, then debit it when the tips are paid out through payroll or as cash. Recording tips as sales overstates revenue and, in most states, inflates the base you are calculating sales tax on.

Sales tax collected works the same way and gets the same treatment. It rides into your bank account inside the card batch, it appears in the deposit, and it is not yours. Credit Sales Tax Payable in the daily entry and clear the account when you remit. Between tips and tax, a restaurant's deposits routinely run 15 to 25 percent above its actual revenue, which is why an operator who tracks sales by watching the bank balance always thinks the business is bigger than the tax return says it is.

QuickBooks Desktop cannot import a bank CSV in any version. There is no hidden setting and no official plugin. The only supported route for transaction data is a Web Connect file, brought in through Banking, then Bank Feeds, then Import Web Connect File. Intuit's own Import Excel and CSV toolkit handles four list types only, customers, vendors, items and chart of accounts, and none of those are transactions.

QuickBooks Online will take a bank CSV, but on terms that restaurant data breaks quickly. The upload wants three columns or four, a single consistent date format, and it caps out at 350 KB and 1,000 rows per file. A busy location's monthly card and delivery activity clears 1,000 rows without effort, and the 350 KB ceiling applies to .qbo, .qfx and .ofx uploads too. Converting to .qbo gives you a file both editions accept and skips the column mapping argument entirely.

Meanwhile the platforms you actually need data from all export CSV and none of them export .qbo. Toast, Square and Clover give you CSV. DoorDash, Uber Eats and Grubhub give you CSV. Your bank gives you CSV, and if it ever gave you a .qbo, that file is increasingly likely to be rejected over its institution identifiers rather than its transactions. Converting the CSV yourself sidesteps all of it.

Build it so prime cost falls out of the P&L without a spreadsheet. Split revenue into food, beverage broken out by beer, wine and liquor, and other income such as merchandise or catering. Split cost of goods sold along the same lines so each category's cost sits directly under its revenue. Put all labor together: hourly wages, salaried management, payroll taxes and benefits.

Prime cost is cost of goods sold plus total labor, expressed as a percentage of sales, and it is the number restaurant operators actually run the business on. The usual benchmark is around 60 percent, with quick service closer to 55 to 60 percent and full service closer to 60 to 65 percent. Fine dining often runs higher because both ingredients and skilled staff cost more. Food cost alone tells you half the story, since a kitchen can hold a great food cost while labor quietly runs away. All of which depends on the top line being real, which brings the whole thing back to booking gross sales rather than deposits.

Keep the operating expenses below prime cost in the same order every month: occupancy, utilities, marketing including the delivery commissions, repairs and maintenance, smallwares, and credit card processing. Processing fees deserve their own account rather than being buried in bank charges, because on a card heavy restaurant they are large enough to manage.

Yes, for the general ledger, provided you let the POS stay the system of record for sales detail. QuickBooks handles the chart of accounts, the P&L, payroll, vendor bills and the balance sheet well. It is not a menu costing system, an inventory counting system or a scheduling system, and trying to make it one is where restaurant books usually go wrong.

The clean division of labor is this: the POS owns tickets, menu mix and hourly sales; your inventory tool owns theoretical food cost and counts; QuickBooks owns the financial statements. The bridge between them is a summary sales entry and a set of clearing accounts, plus a reliable way to get the bank and platform CSVs in without typing them. That last piece is what this converter does.

Export the CSVs you need for the period: the bank account, each card processor if it reports separately, and each delivery marketplace. Drop them in and the converter reads .csv, .txt, .xls, .xlsx and .xlsm, normalizes the dates and amounts, and writes a Web Connect .qbo file you import through Bank Feeds. It also writes XLSX and CSV if you want a cleaned copy for a working paper or for a spreadsheet you already maintain.

From there the close is the ordinary one. Post the daily or weekly sales entries. Match each deposit to its clearing account. Review the clearing balances for the one to two day float they should be carrying. Enter vendor bills, run the prime cost report, and reconcile. What the converter removes is the part that used to take an afternoon, which is getting several hundred bank and platform lines into QuickBooks accurately in the first place.

Restaurants grow by opening rooms, and the second location is where the bookkeeping usually breaks. Use class or location tracking so each site has its own P&L while the entity consolidates. Class tracking is available in every QuickBooks Desktop edition; in QuickBooks Online it starts at Plus. Give each location its own bank account and its own clearing accounts, because merged clearing accounts across sites are effectively impossible to reconcile once the volume rises.

Batch conversion matters more at this point too. Two or three locations means two or three bank CSVs plus the same number of delivery reports each period, and converting them one at a time stops being reasonable. Handling a folder of files in one pass is the difference between a close that fits in a morning and one that eats two days.

Not directly as sales. QuickBooks Online can import a bank style CSV of transactions into a bank account, but a POS daily sales summary is not that shape: it contains revenue categories, tax, tips and payment types that need to land in different accounts. That belongs in a journal entry, not a bank import. What you convert to .qbo is the money side, the bank and card and marketplace activity, so the deposits and fees come in cleanly and the sales entry has something to clear against.

The converter works with the CSV exports those systems produce, since it reads the file rather than connecting to the platform. There is no integration to configure and no credentials to share. Export the report, convert it, import it. The same is true of the delivery marketplaces and of any bank whose download is a CSV, which by now is essentially all of them.