The month-end re-key: a working week lost to a job a machine could do
Ambassadorby Vaughn Botha · 8 min readEvery UK bookkeeping practice runs the same quiet ritual at month end. You export a trial balance from Xero, QuickBooks or Sage. You open your own spreadsheet pack. Then you read a number off one screen and type it into the other, line by line, until the pack is built. It is slow. It feels like care. And it is, increasingly, the wrong way to do it.
This report follows that one figure, from the ledger to the finished pack. It counts what the typing really costs, shows why the tax rules are turning against it, and sets out where a machine genuinely helps and where it does not.1
A third of a UK bookkeeper’s time goes to manual work — and the biggest single piece is rebuilding the same pack, by hand, every month.
The same hour, lost every month
It is the 6th. You run the trial balance. You open the pack. You start typing.
This is not a small job hiding in a corner. It is the job. And the people who do it know how it feels.
“At month end it’s just a really painful and slow process.”— A Sage user, Sage Community forum
Here is the twist. Many bookkeepers type the numbers in on purpose. Going line by line is how they catch a figure that has been coded wrong. It feels like care, and some of it is.
“The copy typing means that we look at each line as it is entered and helps us recognise ‘funnies’.”— A practitioner, AccountingWEB forum
So the job is really two jobs stuck together. Moving the numbers, which is dull and the same every month. And judging the numbers, which is what a good bookkeeper is paid for. Pull them apart and most of the pain goes away.
Why those hours hurt more than they look
The fee for a monthly pack sits at about £150 a client.2 The maths is simple and a little brutal.
There is a second cost nobody invoices: time. Numbers are only useful while they are fresh. Good practices get the pack out by the fifth working day; most should land it within seven to ten. After three or four weeks, one UK guide calls management accounts “practically useless for decision-making”. Every hour of typing is an hour the numbers sit unread.
From April 2026 the rules pull far more clients in — and HMRC has already ruled that copy-and-paste is not a digital link.
The rules are turning against copy and paste
This used to be only about time. Now it is about the law. Under Making Tax Digital, HMRC is blunt.
“HMRC does not consider the use of ‘cut and paste’ or ‘copy and paste’ to select and move information, as a digital link.”— HMRC, VAT Notice 700/22
The soft-landing period ended on 1 April 2021.3 And the scope is about to widen sharply: Making Tax Digital for Income Tax phases in from April 2026, pulling in more clients each year, each filing quarterly.
The manual method isn’t just slow any more. As the rules widen, it quietly becomes the non-compliant one.
“Typing each line is how we catch a figure that’s been coded wrong.”
It feels like care, and some of it is. Reading every number as you key it does catch the odd “funny” — so the manual re-key gets defended as a quality check, not just data entry.
A real reason practitioners give — AccountingWEB forum.
The careful habit is now the non-compliant one.
HMRC says copy-and-paste isn’t a digital link, so the hand transfer is the very part that breaks the rule. A workflow flags every line that moved more than 10 to 15%, which catches the same “funnies” faster. And it tends to turn up errors the manual pass hid. You still review. You just review the flags, after a clean digital pull.
HMRC VAT Notice 700/22; month-end variance practice.
MTD pulls in more clients. The bigger shift is how often each one files.
Today, many of these clients file once a year. Under Making Tax Digital, each one files five times: four quarterly updates plus a final declaration.
How practices cope — and where each fix breaks
Practitioners are not naive about any of this. They have coping strategies. Each one solves part of the chain and breaks somewhere predictable.
| Approach | What it fixes | Where it breaks | MTD-safe? |
|---|---|---|---|
| Smarter spreadsheets | Auto-calculates the pack once the trial balance is in | You still type the figures in; the formulas break when a client adds a code | ✗ Nomanual entry breaks the digital link |
| Accounts-production software | Imports the trial balance automatically | Builds statutory accounts, not your bespoke pack; custom charts map poorly | ~ Partialimports digitally, but isn’t your pack |
| The system’s own reports | No export needed | No custom board pack; “no automated KPI calculations” | ✓ Yescompliant, but no bespoke pack |
| Offshore the typing | Cheaper hours (£8–25 vs £30–90) | Scales the cost, doesn’t remove it; a manual transfer is still manual | ✗ Noa manual transfer is still manual, and the liability stays yours |
Split the job in two
The goal is not speed. It is separation. Hand the dull, repeatable half to a machine. Keep the half that needs a human brain. As one UK guide puts it, the close “should be a review and reconciliation exercise, not a data entry marathon.”
A workflow pulls the trial balance through the system’s own connection, so there is no copy and no paste and the digital link stays intact. It maps each code to the right line using the fixed code ranges that Xero and Sage already use, so the mapping is stable, not a fragile spreadsheet formula.
It works out the changes against last month and budget, flags anything that moved more than 10–15%, and drafts a first cut of the commentary. Then it stops. Nothing reaches a client until a human has reviewed and signed it off.
What is the re-key costing your practice?
Move the sliders to match your practice. The numbers update as you go.
Median UK rate £33.36/hour (6 Figure Bookkeeper Pricing Report, Jan 2026, 131 practices); a working week is 37.5 hours. These are the manual-build hours. Read the figure as recoverable capacity — time a workflow can hand back, not guaranteed cash. The review stays with you, by design.
Picture a typical eight-person practice with around forty clients, fifteen on a monthly pack. That is roughly 30 hours a month of building by hand — close to nine working weeks a year. Move it to a machine and the practice does not shrink; it does more, without the late nights.
Solve it with n8n
I’m an n8n Ambassador.The tool I build in is called n8n. The simplest way to picture it: a row of small jobs, wired together, that run on their own. Each job does one thing and hands the result to the next. I set the rules. You stay in charge of what goes out.
Here is the same month-end pack, built as a machine. It pulls your trial balance straight through the software’s own connection, so there is no export and no paste, and the digital link the rules now demand stays intact the whole way. It maps each code to the right line using the fixed code ranges Xero and Sage already publish. It works out what moved against last month and budget, and writes a first draft of the notes. Then it stops and waits for you.
That pause is the whole point. The machine never sends anything to a client on its own. It does the typing and the sums, then puts a finished draft in front of you to check and sign. You keep the judgement. It takes the grind.
Built to sit safely next to a live client ledger.
- Authorised access, not scraped exports.The workflow connects through Xero, Sage or QuickBooks’ own authorised connection — the same OAuth login you already use. No passwords are stored.
- Runs in your environment. The flow lives on your own n8n instance, under your control — not on some shared server of mine.
- No long-term data retention.Figures pass through to build the pack and are not kept afterwards. The machine moves data; it doesn’t hoard it.
- UK-GDPR posture, your sign-off. Processing is set up to fit UK data rules, and nothing reaches a client until you review and approve it.
I build and hand over the workflow; you stay the data controller for your clients throughout.
The monthly pack: by hand, then with the machine
The machine flags the lines that moved and says why. You still read them, still catch the thing that does not look right, still put your name to it. The difference is you’re reviewing in minutes instead of rebuilding from a blank page for hours.
Three things to do about it
The build. A one-off engagement: I study your stack, build the workflow, and hand it over documented. You own it.
Where it runs. On your own n8n instance, under your control. I can help you stand one up, or work with the one you already have.
Optional upkeep. A light retainer if you want me to maintain and extend it as your clients or the rules change — never required.
Prove it small. We pilot on three clients before you trust it on thirty.
n8n AmbassadorIf your month-end looks like this
I’m Vaughn — I wrote this report and I build the workflow behind it. NearTrigger studies one business grind at a time, then ships the machine that fixes it. If the re-key is eating your month-end, I’ll take a free look at yours and show you where a workflow earns its place — and where it doesn’t. No client data shared, no obligation.
Book a free month-end teardown →About this research
This report is built from live UK sources: practitioner forums (AccountingWEB, Sage Community), HMRC and GOV.UK guidance on Making Tax Digital, professional-body guidance (ICB, ACCA, ICAEW), software documentation, and UK accounting blogs and templates. Every number comes from a named source. Estimates are labelled as estimates. The example practice is illustrative, a stand-in for the patterns in the research, not a real client. Anything I could not verify, I left out.
1. Built from 54 UK sources retrieved while writing, June 2026.
2. Median monthly fee £150.50 and median hourly rate £33.36: The 6 Figure Bookkeeper Pricing Report, January 2026 (131 practices).
3. HMRC VAT Notice 700/22; soft-landing end date per ACCA.
Vaughn Botha is an n8n Ambassador based in KwaZulu-Natal, South Africa, and the author of NearTrigger Field Research. This is Field Report No.01.
NearTrigger Field Research No.01 · Published June 2026 · neartrigger.co.za