Why clients never send receipts (and what actually fixes it)

Every accounting firm sends the same email every spring: *"Can you tell us what this $84 charge from Home Depot last June was for?"* And every spring the answers come back identical: "no idea," "probably materials?", "can we just call it supplies?"

We run an accounting firm, and we've sent thousands of those emails. Here's what we finally admitted to ourselves: the receipt problem is not a discipline problem. Your clients aren't lazy and neither are you. It's a timing problem. Nobody can answer questions about June in March. Everyone can answer them *in June* — usually in the ten seconds after the purchase, while they're still holding the receipt.

Every system that fails, fails because it asks for the document long after that moment has passed. Every system that works, works because it captures the document inside it.

Why the standard rituals fail

• The shoebox (and its digital cousin, the "receipts" folder) postpones the work to a mythical future afternoon that arrives, if ever, at tax time — after the ink has faded and the context is gone. The shoebox was never a system; it's the absence of one.
• The monthly "please send me everything" email asks the client to do a batch chore with no definition of done. "Everything" is not a list. Nobody finishes a task that can't be finished.
• The scanning-app Sunday session front-loads friction: install an app, log in, photograph forty crumpled receipts in a sitting. The people disciplined enough to do this weren't your problem in the first place.
• Chasing by human — a bookkeeper writing personal nag emails — works, which is why firms do it, and burns hours of skilled labor on secretarial work. It also quietly caps how many clients one bookkeeper can carry.

The three mechanics that actually work

After years of trying everything on real clients, the systems that get documents into books share three properties:

1. Capture at the moment, with zero new habits. The winning motion is: snap a photo, send it the way you already send everything else. Forward it to an email address. Drop it in the portal from your phone. No app to install, no login to remember, no categorizing at capture time — the *only* job at the moment of purchase is getting the pixels out of the physical world. Every additional step at capture time cuts compliance roughly in half.

2. Chase the list, not the concept. "Send me your receipts" gets ignored; "these six transactions are missing documentation" gets answered — because it's specific, finite, and mostly one-tap from a phone. The chase also shouldn't be a person. It should be a system that knows exactly which transactions lack support, asks for precisely those, says thank you, and quietly asks again in a few days about whatever's still missing. Machines are better at polite persistence than humans are, and no relationship capital gets spent on nagging.

3. Land the document on the transaction. A receipt in a folder is storage; a receipt attached to the ledger entry it supports is bookkeeping. If the intake pipeline doesn't end with the document connected to the transaction — matched, or held for a human to confirm the match — you've just built a tidier shoebox. This is also what makes tax season quiet: the answer to "what was this $84?" is stapled to the $84.

Any stack can implement a version of this: a dedicated intake email per client, a standing weekly report of undocumented transactions, and someone (or something) that sends the missing-list. The pieces matter more than the brand names.

Where Woven stands

Disclosure: Woven is our product, and this problem is one of the reasons it exists. In Woven, document intake is a function of the ledger, not a companion app: every client's books get their own inbound email address, and photos or PDFs sent there — or dropped through the client portal from a phone — land in that client's intake queue. The AI reads each document and proposes the match or the entry with the document attached; a person approves it, always. And the receipt chase runs automatically: clients get a periodic message listing exactly which transactions still need documentation, and the list shrinks as things arrive, without a bookkeeper writing a single reminder.

It was built by the owner of an accounting firm whose own clients supplied the shoeboxes that motivated it.

The short version

• The receipt problem is a timing problem. Capture wins at the moment of purchase or not at all.
• Zero-friction inbound beats every app: photo → email or portal, done.
• Chase a specific list of undocumented transactions, automatically — never "everything," never by hand.
• A document that isn't attached to its transaction is still a shoebox, just alphabetized.

If your firm is still hand-writing the chase emails — or you're the owner with the glovebox full of receipts — [book a demo](/demo) and we'll show you what intake inside the books looks like.