Hobfolk
Money · case study

Our turnover and the bank
were reconciled to the penny.

A set of first-year statutory accounts drew a query over a gap between two turnover figures. Rather than guess from the bank statement, the actual payment ledger was pulled — thousands of individual entries — and our side of the gap was bridged to the penny, line by line.

Start a conversationSee the whole crew
You go home. It doesn't.Every promise gets a row and a due dateEvery enquiry answered, not just loggedA drawing before the quoteUnverifiable is never a passA desk per department, one address eachNothing is sent that a human cannot read backThe folder exists before you price itYou go home. It doesn't.Every promise gets a row and a due dateEvery enquiry answered, not just loggedA drawing before the quoteUnverifiable is never a passA desk per department, one address eachNothing is sent that a human cannot read backThe folder exists before you price it
The story

A turnover gap, bridged from the ledger, not the bank statement

The bank statement tells you what arrived. It does not tell you what was taken before it arrived — and that difference is where a real gap hides.

An accountant preparing a first set of statutory accounts queried a gap of several thousand pounds between two turnover figures. The instinct in that situation is usually to work backwards from the bank statement, which is exactly the wrong instinct when the business sells through a marketplace — the bank only ever sees what's left after the platform has already taken its cut.

Instead, the marketplace's own payment ledger was pulled directly: several thousand individual entries, itemising platform fees, advertising spend, shipping labels bought against the account, refunds and clawbacks, all deducted before the money ever reached the bank. Lined up against the two figures, that itemised list accounted for almost the entire gap on its own.

What was left after that was a much smaller, odd remainder — and the likeliest cause was a transfer between accounts that is not income at all, so the accountant was asked for his line items to settle it rather than have it guessed. A second error came out of the same pass: marketplace fees and ad spend had been booked against an estimate rather than the real recorded figure, a few hundred pounds out, and were corrected in the draft accounts rather than found after filing.

Thousands of entries, read once, not guessed at
Thousands of entries, read once, not guessed at

Sourced from Hobfolk's own operational record, dated and on file — see the other cases. Client identity is never named; the numbers and the root cause are not softened.

What changed

The permanent fix, not just the patch

01

Reconciliation runs from the platform's own ledger

Never from bank-statement inference — the bank only shows what's left after deductions, not the deductions themselves.
02

Every at-source deduction is itemised

Platform fees, advertising, shipping labels bought against the account, refunds and clawbacks are each accounted for by name, not lumped into one unexplained gap.
03

The remainder is chased, not rounded off

Not accepted at “close enough” — the remainder is chased until it has an explanation, and the accountant is asked for the line items that settle it.
04

Estimated figures are replaced before filing, not after

A fees figure booked against an estimate rather than the real recorded number was found and corrected in the same pass, in the draft accounts.
05

The accountant gets a working paper, not just an answer

The full reconciliation table, not a verbal reassurance that the numbers are fine.
06

The same check now runs on every accounts query

Not treated as a one-off fire drill — the ledger-first method is the standing approach now.
How it unfolded

The sequence, in order

  1. The accountant raises a query

    A gap between two turnover figures on a first set of statutory accounts.

  2. The ledger is pulled, not the bank statement

    The marketplace's own payment ledger — thousands of individual entries — rather than working backwards from what landed in the bank.

  3. Deductions are itemised by name

    Platform fees, ad spend, shipping labels, refunds and clawbacks each accounted for separately.

  4. The gap narrows to a small remainder

    Almost the whole difference is explained by the itemised list alone.

  5. The remainder is narrowed to one likely transaction

    A transfer between accounts that is not income; the accountant's line items are asked for to confirm it.

  6. A second, unrelated error is caught in the same pass

    An estimated fees figure, corrected to the real recorded number in the draft accounts.

The limits

What this desk will not do

A case study is not a claim of no limits. The same boundary applies here as on the department page this one came from.

Where this desk stops

  • It never contacts a creditor. Money going out is yours, personally and always. There is no automated desk on that side of the ledger and there never will be — a system that chases money in must not be able to speak to someone you owe.
  • It does not move money. It prepares payments; you release them.
  • It does not negotiate or agree settlement. It asks when an invoice will be paid. That is the whole remit.
Questions

Asked before, answered here

Why not just check the bank statement first?
Because for any business selling through a marketplace, the bank only ever shows what's left after the platform's own deductions — fees, ads, shipping, refunds — have already been taken. Working from the bank statement first means reconciling against a number that was never the full picture to begin with.
How do you know the reconciliation is actually complete?
By reconciling to zero, not to “close enough.” Any remainder is chased to a specific, named cause, with the other side's line items asked for where they are needed, rather than written off as immaterial.
Does this replace our accountant?
No. The accountant owns the statutory filing and its presentation; this desk prepares the reconciled figures and the working paper behind them, so the accountant's query gets an itemised answer instead of a guess.
What if the second error hadn't been caught in the same pass?
It would very likely have gone into the filed return uncorrected — which is the actual case for reconciling from source data rather than accepting the first plausible explanation for a gap and stopping there.
The department

This case belongs to the Money desk

Money

Invoices out, cash in

Invoices raised, books reconciled, cash position current, and a separate desk that does the asking.

Open the department Case studies

Every case, in one place

Real, dated things that went wrong in the businesses this office already runs, and the fix each one produced.

See all the cases
Two more cases

More from the record

Case study

Reference photo to tested mechanism

A design that had to move correctly, not just look correct

Read the caseCase study

Every listing live, none of them true

A hunch, checked properly, turned out to be exactly right

Read the case

Could you explain your own last accountant's query from source data today?

Tell us where the gap is, and we'll tell you whether it's a ledger problem or a bookkeeping one.

Start a conversation Read the method