What is messy document intake?
Messy document intake is everything that happens before bookkeeping can start: receiving receipts, invoices, statements, phone photos, and bundled PDFs from clients in no particular order, then sorting, splitting, reading, and preparing them into records an accounting system will accept. It is the prep work, not the bookkeeping.
Messy document intake, defined
Messy document intake is the pile before the books. Clients do not send clean inputs; they send what they have. A combined PDF. A shuffled scan. Phone photos. Receipts tucked inside invoices. A file named IMG_4471. It is technically submitted, but it is not ready for anything yet. Intake is the work of turning that into sorted, classified, accounting-ready records.
Why it is the real time sink
From the outside it looks like bookkeeping. It is not. Opening every file, working out what belongs with what, separating documents, reading the amounts, typing them in, and then checking what you typed, almost all of that happens before a single entry is posted. That is where the hours actually go, and where avoidable errors start. Bookkeeping is not the problem. The prep in front of it is.
The everyday form is an unstructured dump, so drop it in as it is
In practice, messy intake usually arrives as an unstructured dump from a client: a folder or end-of-month email with a forty-page PDF, a dozen phone photos, a bank statement, and a contract, none named usefully and several documents stuffed into one file. You do not pre-sort, rename, or split any of it. You add the business, open Cost and Spending, and drop the files in as they came. The standard advice, train your clients to send tidy files, pushes the work onto people who will not keep it up and back onto you when they do not. untxt. is built to take the dump as it is, so you stop policing how clients upload.
Intake is a layer, not a "scan to spreadsheet" trick
The common framing is OCR to spreadsheet: read the text, export a CSV, done. That handles the easiest case, one clean invoice, and then leaves you babysitting the export: importing it, fixing what broke, mapping it by hand. That export-and-cleanup loop is the part bookkeepers say they hate. Real intake is messier than one page anyway, because the hard work is understanding the pile. So untxt. sits as the intake layer in front of QuickBooks and Xero. It sorts the batch, splits bundled files, classifies each document, extracts the fields and supported line items, maps to your chart of accounts, flags what it is unsure about, and sends the reviewed data straight into QuickBooks or Xero, with no export to babysit. If you do want the files, you can still download CSV, TXT, JSON, or DOC, one format or all at once.
What intake still leaves to human review
Intake does not make review vanish. It makes review narrower. Handwriting, signature scrawl, and genuinely ambiguous documents still need a person, and untxt. flags those for you to confirm. Entries post only after you have seen them. The point is to stop you rebuilding the pile from scratch so you can spend your attention on the exceptions that actually need judgment.
Who runs into messy document intake
Solo bookkeepers, multi-client practices, and businesses keeping their own books all hit the same wall: imperfect documents arriving faster than anyone can hand-sort them. If your month begins by triaging a folder of mixed files, that is messy document intake, and it is the specific thing untxt. was built to take off your desk.