2026-09-15

Inside one fix: a market vendor's paper purchase log became an app she actually uses

She sells food at a local market. She's just starting out, and until recently, every purchase she made for the business went into the same place: a notebook, written by hand, whenever she remembered to do it.

That's not a knock on her. It's how most owner-operated businesses run in year one — and for a lot of them, it's how things stay for years after that.

Paper and pen worked, until the market didn't

The hard part of her buying pattern isn't the volume. It's the timing. She's not placing orders from a supplier on a schedule. She's walking a market, seeing a good price on something she needs, and buying it on the spot — sometimes from three different sellers in twenty minutes. There's no invoice email, no receipt that lands in an inbox, nothing that gets captured automatically. If she doesn't write it down right then, the only record is her memory.

A notebook handles that fine for a while. The problem shows up later, at reconciliation — when she's trying to work out what she actually spent, on what, and whether it matches what's in her bank account. Reading back her own handwriting from two weeks ago, filling in gaps from memory, is where the real time and the real errors creep in.

What we actually diagnosed

The instinct when someone says "I'm tracking purchases on paper" is to hand them accounting software and call it solved. We didn't do that here, because that's not actually the problem.

The problem wasn't that she lacked a system. It's that any system had to work at the exact moment she was standing at a market stall, coins in hand, with a seller waiting. Software that assumes you're sitting at a desk — categorizing, tagging, matching to a chart of accounts — doesn't survive contact with that moment. She'd stop using it within a week, and be back to paper.

That's the actual shape of the problem: capture has to happen in under a few seconds, in the field, with no setup. Everything else — categorizing, totaling, reconciling — can happen later, when she has time to sit down.

Why a mini-app and Google Sheets, not a bookkeeping platform

So the fix isn't a platform. It's a small, purpose-built web app that does exactly one thing well: let her log a purchase — what she bought, from whom, how much — in a few taps, the moment she makes it. No account setup to explain, no chart of accounts to configure, nothing to learn beyond "tap, enter the amount, done."

Every entry she logs writes straight into a Google Sheet behind the scenes. That's a deliberate choice, not a shortcut. She already knows what a spreadsheet is. She doesn't need a new login, a new export step, or a new tool to trust — she can open the sheet any time and see exactly what's in it, in a format she already understands. The reconciling — matching what she logged against her bank statement, spotting anything odd — happens in that same sheet, on her own time, not mid-market.

This is the same idea behind choosing the right fix format for a business generally: match the fix to what someone can actually use today, not what's technically the most complete solution.

How it works, purchase by purchase

In practice, it looks like this: she sees something cheap she needs, buys it, and opens the app for a few seconds before she's even walked to the next stall. Item, seller, amount. Done. No notebook, no trying to remember it that evening.

Later — at home, or whenever she has a spare ten minutes — the Google Sheet has every purchase already sitting there, timestamped, ready to check against her bank account instead of against her memory.

What she's saying now that she's using it

The app has been in her hands for a short time, and she's actually using it — logging real purchases at real markets, not just trying it once and going back to the notebook. She's also giving us feedback as she goes: what's fast enough, what's still fiddly, what she wishes it did. That back-and-forth is shaping what gets built next.

It's early. There hasn't been a full reconciliation cycle yet, so we're not going to claim a number of hours saved that we haven't actually measured. What we can say honestly: the notebook has stopped being where her purchases live, and she's choosing to open the app instead, market day after market day. For a fix built to survive exactly one moment — her, at a stall, with a seller waiting — that's the test that matters.

This is what diagnosis-before-automation looks like in practice, for a business with three days a week to run a stall and zero time to learn new software. Book a call if a piece of your own business still lives in a notebook or a memory, and you want to know what a fix matched to your day actually looks like.