After-Hours Bookkeeping: Half of Self-Employed Accounting in Germany Happens Outside Business Hours

When does a country do its bookkeeping? For employees the answer is simple: during working hours, because there is an accounting department for it. The self-employed have none. Their bookkeeping competes with the actual business for the same hours, and it loses that competition almost every time, until the client work is done and the laptop opens one more time.

This analysis is based on about 5,000 active bookkeeping hours of nearly 1,000 self-employed people and small businesses in Germany, covering January 2025 to June 2026. An active bookkeeping hour is an hour in which a business performed at least one bookkeeping action: wrote an invoice, uploaded a receipt, recorded a transaction by hand. So we do not estimate when the self-employed do their books; we count it.

Every second hour falls outside business hours

Apply a generous definition of business hours (Monday to Friday, 8am to 6pm) and half of all active bookkeeping hours fall outside it (about 50 percent). Nearly three in ten hours start after 6pm, about one in eight falls in the night between 10pm and 6am, and one in five lands on the weekend.

This is not an average produced by a few night owls. Looking only at regularly active businesses, the typical one does about 44 percent of its bookkeeping outside business hours. For four in ten, the majority of their bookkeeping happens after hours, and for more than eight in ten it is at least a quarter. Nearly nine in ten regularly active businesses have done bookkeeping after 6pm at least once.

Germans have a word for the moment the working day ends: Feierabend. For the self-employed, that is often when the books open.

9pm is busier than 9am

Download CSV

The daily curve tells the story in detail. It rises through the morning, peaks in the afternoon around 3pm, and then does not fall off: it forms a long evening plateau. At 9pm the share is higher than at 9am (5.6 versus 4.7 percent), and even at 10pm nearly as much bookkeeping happens as at the start of the classic working day. This holds even when looking at weekdays only.

The working day of the self-employed does not end; it relocates. First comes the business, then come the receipts.

Sunday is the secret bookkeeping day

Download CSV

The weekend shows the clearest pattern. Saturday is by far the quietest day of the week, while Sunday is almost a normal working day: it accounts for 12 percent of all active bookkeeping hours, one and a half times as much as Saturday. Nearly two in three regularly active businesses have done their books on a Sunday at least once. Saturday belongs to life; Sunday belongs to preparing for the week.

And the week starts with a cleanup: Monday is the strongest bookkeeping day at 18.5 percent, and invoices, too, are written most often on Mondays. The cliché of Friday as the big invoicing day finds no support in the data; Friday is, together with Wednesday, the quietest weekday.

Invoices follow the turn of the month

Download CSV

The week has one rhythm; the month has a second, and it swings hardest in invoicing. The strongest invoicing day of the month is the first working day: 1.6 times as many businesses issue invoices on it as on an average working day. Right behind it comes the last working day of the month, at 1.5 times. Between the two lies a deep trough: from the 8th to the 22nd, invoicing activity runs about a fifth below the average day. At the turn of the month, nearly twice as much is happening as in mid-month.

The pattern fits two billing logics that meet at the month boundary: those who bill ongoing work draw the line under the month on its last working day; those who bill the previous month do it on the first. What stands out is how specific the effect is. Receipts are processed almost evenly across the month and show no comparable arc. Bookkeeping as a whole follows the week; invoicing follows the month.

The 10th of the month sets the country's beat

Download CSV

One rhythm, however, is set by the calendar. The advance VAT return (Umsatzsteuervoranmeldung) is due on the 10th of the following month, and the deadline works. The 10th is by far the busiest submission day of the month (about 13 percent of all returns), about one in five returns arrives in the last 48 hours before its due date, and after the 10th activity drops off abruptly. On top of this sits a quarterly beat: in the months after a quarter ends (January, April, July, and October), 2.4 times as many returns come in as in the other months: that is when the quarterly filers meet the monthly ones. About half of all returns are verifiably submitted within the standard deadline; the rest is a mix of extended deadlines (Dauerfristverlängerung), corrections, and stragglers that the data cannot cleanly separate.

Why it is this way, and what it means

Being self-employed means there is no one to delegate the receipts to. Bookkeeping is the invisible second shift of self-employment, and it gets done when the first shift, the actual business, takes a break. One can read that as flexibility or as work without boundaries; what the data establishes is that this shift exists, and how large it is.

The pattern is remarkably stable. It shows up in every single action type, from writing invoices to uploading receipts, at nearly the same strength in both user groups we analyzed, and in every quarter of the period. Even the questions people ask our bookkeeping assistant follow the same beat: about half of them, too, arrive outside business hours.

The data's limits deserve naming: whoever runs their books on a digital platform belongs to the digitally working half of self-employment. Whether the after-hours share would be higher or lower among the self-employed with a tax advisor and a paper binder, our data cannot say. But for the digitally working, the conclusion stands: bookkeeping has no fixed slot in the calendar. It fills the margins.

This analysis will be updated annually at this address; the changelog at the bottom of the page documents every revision. All analyses in this section: Data & Research.

Methodology

The basis is about 5,000 active bookkeeping hours of nearly 1,000 self-employed people and small businesses in Germany, January 2025 to June 2026. An active bookkeeping hour is a calendar hour in which a business performed at least one of three bookkeeping actions: created an invoice, uploaded a receipt (via upload or in chat; documents pulled automatically from bank feeds do not count), or recorded a transaction manually. Each business counts at most once per hour, so bulk uploads and automatically generated series (such as recurring invoices) do not distort the picture. All times are German local time (Europe/Berlin). Business hours are defined as Monday to Friday, 8am to 6pm. Internal and test accounts are excluded. The pattern is robust: it appears in every action type individually (44 to 56 percent outside business hours), at nearly the same strength in both sub-populations of our user base, and in all six quarters of the period (48 to 58 percent). Figures about the "typical business" refer to businesses with at least eight active hours in the period.

The monthly invoicing rhythm counts businesses per calendar day (each business counts at most once per day) and is normalized per occurrence of the calendar day, since the 31st occurs less often than the 10th; working days are Monday to Friday, without adjusting for public holidays. The turn-of-the-month pattern appears in both halves of the period. The 18-month series does not yet allow serious claims about seasonal effects across the year (such as a January high); we will add them once the series is long enough.

The advance-VAT-return analysis rests on more than 400 electronically submitted returns from more than 100 businesses since mid-2024. The due date is the 10th of the following month, or the next working day; a Dauerfristverlängerung shifts the month, not the day. Submissions after the due date cannot be unambiguously separated into stragglers, corrections, and extended deadlines; we therefore only report shares that are verifiably on time.

All values are shares; no published figure is based on fewer than 40 businesses. We do not publish absolute user numbers. The sample represents digitally working self-employed people. How we work with data in general is documented in Methodology & data basis.

About this data

Norman is an accounting platform for self-employed people and small businesses in Germany; the anonymized usage data analyzed here originates there. Charts and figures on this page may be used freely with attribution to "Norman". Press inquiries and data requests: press@norman.finance.