Keeping records when you hate paperwork

You know you should be writing this down. You've known for four months. There's a folder on your phone called "receipts" with eleven blurry photos in it, a bank account with your rent and your fabric supplier in the same list, and a growing suspicion that January is going to be unpleasant.

The whole system is four columns and ten minutes a week. Date, what it was, money in, money out. A fifth column for which account it went through if you're mixing cash and bank, which you probably are. That's it. You don't need software, you don't need a bookkeeper, and you don't need the $19 a month for a plan with CSV export until your order count makes exporting faster than typing.

What you do need is to record it at the time. Everything painful about bookkeeping comes from doing it later.

The reason the shoebox fails is arithmetic, not willpower

Recording a transaction while it's happening takes about eight seconds. You know what it was, you know why, the receipt is in your hand.

Reconstructing it in March takes about a minute if you're lucky and ten if you're not. You're squinting at "SQ *KRAFTHOUSE 442" on a card statement, trying to remember whether that was card stock for orders or a birthday present. Your bank app only shows you so far back. The supplier's site has archived the order. The market organiser who took $45 in cash gave you a paper slip that no longer exists.

Call it eight times the effort, minimum, and the multiplier grows the longer you leave it. That's the entire argument. Not discipline, not being organised, not being the kind of person who has a system. Just the fact that a job costing eight seconds now costs several minutes later, and you have hundreds of them.

The receipt you'll actually lose is the one from the supermarket at 10pm when you ran out of boxes, and packaging is exactly the category anyone reviewing your numbers looks at first.

What your records have to be able to answer

Not "be tidy". Answer questions. There are four that matter, and any system that answers them is good enough.

What came in? Every sale, gross, before fees. Not what landed in your account. The difference between those two numbers is fees, and if you only record the net you'll never see what the fees cost you.

What went out? Materials, packaging, postage, the subscription, the market stall, the sample you sent a creator, the app you pay for annually and forgot about.

Who owes what? Orders taken and not paid. Invoices sent and not settled. Supplier bills due. This is the column that stops you from thinking you're rich.

What's sitting in boxes? Stock you paid for and haven't sold. It's real money and it isn't in your bank account, which makes it the number people most often forget exists.

If you can answer those four on any given Tuesday, you have adequate books. If you can't, adding software won't help, because software just makes the same four questions prettier.

One row per thing, and the row looks like this

Here's a real week for a seller doing about twenty orders a month, half online and half from a Saturday market. Nothing clever. One line per event.

Date	What	Category	In	Out	Where
2 Aug	Order 1043, 2 prints, Rivera	Sales	70.00		Bank
2 Aug	Postage, 2 tubes	Postage		11.90	Card
3 Aug	50 mailing tubes	Packaging		62.00	Card
4 Aug	Market stall fee, Saturday	Fees		45.00	Cash
4 Aug	Market sales, 6 prints	Sales	210.00		Cash
5 Aug	Ink, two colours	Materials		38.40	Card
6 Aug	Sailo Business	Subscription		49	Card
7 Aug	Order 1044, 1 print, Okafor	Sales	35.00		Bank

Eight rows, one week, maybe ninety seconds of typing spread across seven days.

Two things about that table are doing more work than they look. The Category column is what turns a list into an answer, because at year end you sort by it and the totals are the numbers you actually need. Keep the categories to about eight and never invent a ninth at 11pm. Sales, materials, packaging, postage, fees, subscriptions, equipment, other. That's plenty.

The Where column is what lets you check yourself. At the end of the month, everything marked Bank should add up to what your bank statement says. If it doesn't, something's missing, and finding it now is trivial compared to finding it in March.

Receipts: capture at the counter, sort never

The mistake is planning to file receipts. Nobody files receipts. Do this instead.

Photograph it before you leave the shop. Not in the car. Not tonight. At the counter, while it's flat and legible, because a thermal receipt in a pocket is illegible within a fortnight and blank within a few months.

Then name it, in the photo app or by dropping it into a folder: 2026-08-03 uline tubes 62.00. That's the whole filing system. When someone asks for the receipt for a $62 charge in August, you search "62" and it's there.

Email receipts are easier and people still lose them. Make one folder, one rule: anything with an amount and a supplier name goes in it, on arrival. Don't read it, don't decide, just file it.

How long you have to keep any of this varies by country and by document type, and it's longer than most people assume. What receipts to keep and for how long goes through it properly. Look up your own country's period rather than guessing, because "I thought it was three years" is not an answer anyone accepts.

Ten minutes, same slot, every week

Pick a slot. Tuesday morning with coffee, Sunday evening before the week starts, whatever you'll actually keep. Put it in the calendar as a repeating thing so it exists whether or not you feel like it.

In that ten minutes:

1. Open the bank app and the sheet side by side.
2. Every line in the bank that isn't in the sheet, add it.
3. Every line in the sheet marked Bank that isn't in the bank, chase it. Usually that's a customer who said they'd transfer and didn't.
4. Empty the receipts folder into the sheet.
5. Count the cash tin and write the number down.

Step three is the one that pays for the other four. Weekly checking is how you notice that an order marked paid never actually landed, while the customer still remembers the conversation and you still have the messages. Do it monthly and you're arguing about something from five weeks ago. Do it never and you ship goods you were never paid for, which is far more common than sellers admit to each other.

Cash is the part that breaks

Bank transactions record themselves. Cash doesn't, and cash is where small sellers lose the thread.

Three rules that fix most of it:

One tin, and nothing else in it. Not your wallet. Not "I'll pay myself back later". Cash sales go in, cash purchases come out, and both get written down before the tin closes.

Bank it on a schedule, not when it feels like a lot. Every Monday, or after every market. A deposit on a statement is a record; a wad in a drawer is a memory.

Write the market day down before you drive home. Total taken, stall fee, float you started with, float you ended with. Do it in the car park. By the time you're home you'll have rounded it, and rounding is how a $210 day becomes a $200 day forever.

Cash on delivery adds its own layer, because the money reaches a courier days before it reaches you. Record the sale on the day the parcel goes out, record the remittance separately when it lands, and never treat the two as one event. The gap between "delivered" and "remitted" is where most COD sellers lose track of a few hundred at a time.

Marcus, screen prints, Baltimore

Marcus screen-prints posters and sells them at $35, about 22 a month online plus a Saturday market that does another 6 or 8. Call it $1,000 a month gross in a decent month.

He kept nothing for his first seven months. His view, which is a reasonable one, was that he'd rather spend the time printing.

What that cost him showed up in three places.

He'd been charging $5.95 for shipping on a tube that cost $11.90 to send. He knew postage was "around six dollars" because that's what it cost the first time, on a shorter route, before a rate change. Twenty-two orders a month, roughly $4 short on maybe two-thirds of them, is about $60 a month walking out of the door invisibly. He found it the week he first added a Postage column, because the total looked wrong.

He'd forgotten a $180 annual charge for a design tool that renewed in February. Not a disaster, but it was in the bank and not in his head, and it changed what he thought a month cost him.

And he had two orders from a repeat customer, worth $105, that he'd shipped and never been paid for. Both were "I'll send it tonight" messages in Instagram DMs from a Tuesday. He found them by scrolling back through five weeks of conversations, and he only found them because he was reconstructing the whole period at once.

He now spends about eight minutes on a Sunday evening. The system is a spreadsheet with the six columns above and a phone folder of photographed receipts. It has never taken longer than fifteen minutes, and the $60-a-month postage discovery paid for the rest of his life's bookkeeping in the first month.

When you're already six months behind

Most people reading this aren't starting clean. Recovering is a different job from maintaining, and doing it in the wrong order is what makes people give up halfway.

Do it backwards, and do it in two passes.

Pass one: money in. Export or scroll your bank statements for the period and record every credit only. Ignore expenses entirely for now. This is the number that matters most and it's the easiest to reconstruct, because credits are few and large. If you've been taking card payments, your processor has a full record you can export. If you've been taking transfers, the bank has it.

Pass two: money out, biggest first. Sort by amount, largest to smallest, and work down. The top twenty transactions will be most of your spend. Getting the top twenty right and estimating nothing below your smallest supplier order is far better than starting in January and running out of energy in March. Anything you genuinely can't identify, mark as unidentified and move on. An honest gap beats an invented row.

Then draw a line, start the weekly habit from today, and don't try to be perfect about the part behind you. The point of the recovery pass is a defensible number, not a beautiful ledger.

Don't buy software in month one

Accounting software is worth it when the typing becomes the bottleneck, and typing doesn't become the bottleneck until you're somewhere north of a hundred transactions a month. Before that it's a subscription that adds a login and a learning curve to a job a spreadsheet does.

The order that usually works: spreadsheet, then spreadsheet plus CSV exports from wherever your orders live, then real software when your accountant asks for it or when you hire someone.

Sailo's free plan gives you 7 days of analytics, Pro gives you a year and CSV export at $19 a month, Business gives you three years. Which brings up the limitation worth stating plainly, because it applies to every shop platform and not just this one: your shop knows what came in, and it has no idea what went out. It doesn't know you spent $62 on tubes or $45 on a stall. It can't tell you whether you're profitable, only what you sold. And on the free plan the analytics window is 7 days, which is shorter than the period any tax authority will expect you to be able to account for. Your sheet is the record. The shop is one input to it.

That's also why exporting matters more than it sounds. Export your orders on a schedule and keep the file, whatever platform you're on. Data inside somebody's product is data you're renting.

What this feeds

Records aren't the point. They're the thing that makes three other jobs possible.

They're where your tax return comes from, and reporting profit rather than turnover is the whole game. What to do about tax on online sales covers the shape of that, and the reason people overpay is almost always unrecorded expenses rather than anything clever they missed.

They're how you find out whether the shop actually makes money once packaging, fees and your own hours are counted. Working out if you are actually profitable runs the full arithmetic on one basket, and it needs numbers you can only get from a ledger.

And they're only workable if the shop's money isn't tangled with your own. Separating business and personal money is the prerequisite, honestly, and doing it first makes everything above take half as long.

If you're not sure yet whether any of this applies to you, do you need to register your business sorts out which obligations you're actually under.

Start with the last seven days

Open a spreadsheet right now. Six columns: date, what, category, in, out, where. Fill in the last seven days from your bank app and your messages. It'll take about fifteen minutes and it'll be incomplete, which is fine.

Then put a repeating ten-minute slot in your calendar for the same time next week, and photograph the next receipt you're handed before you put your wallet away.

The backfill can wait. The habit can't, because every week you delay adds another eight-times-harder week to the pile.