A close package has one job: to let a reviewer form a defensible opinion about the numbers quickly. Every part of it either serves that job or is wasting your time.
Most packages fail in one of two directions. Some contain too little — statements with no supporting schedules, which forces the reviewer to open the ledger and rebuild the work. Others contain everything the software could export, which forces the reviewer to work out what matters. Both cost the same thing: your hours.
Here is what a package should contain, and why.
The statements
Balance sheet, income statement and trial balance, with comparatives.
Comparatives are not decoration. A reviewer reads a close package by looking for what changed, and a statement without a prior period makes that impossible. Prior month and prior year to date should both be available.
The income statement should be presented the way the client’s business actually works — by department, class or location if that is how the operation is managed. A single undifferentiated column tells nobody anything.
The reconciliations
Every bank, credit card and loan account, reconciled to a source statement.
Two things make this reviewable rather than decorative:
- The statement itself is attached, not just the reconciliation report. A reconciliation that agrees to a number nobody can verify is an assertion, not evidence.
- Outstanding items are listed and aged. A cheque outstanding for eleven months is a finding. A reconciliation that nets to zero while carrying it silently is hiding one.
Merchant and payment-processor clearing accounts belong here too. They are where errors accumulate quietly, because the balance is small and nobody looks.
The supporting schedules
Anything on the balance sheet that is not a reconciled cash account needs a schedule behind it. Prepaid expenses, accrued liabilities, fixed assets and amortization, deferred revenue, loans with amortization tables, intercompany balances.
The rule of thumb: a reviewer should be able to agree every balance sheet line to something without opening the ledger. If they have to go digging, the package has moved work rather than done it.
The agings
Accounts payable and accounts receivable, aged, tied to the general ledger.
The tie-out matters. A subledger that does not agree to the control account is one of the most common quiet errors in bookkeeping, and it compounds. The package should state that the tie has been performed, not leave the reviewer to check.
Commentary on the outliers is worth more than the report itself. A receivable at 120 days is a fact; a note saying the client has stopped responding to invoices is information your firm can act on.
The open items list
One consolidated list of everything unresolved, prioritized, with the document trail attached.
This is the part most packages get wrong, and it is the part that most affects your hours.
The failure mode is the suspense account. An item cannot be identified, so it is posted somewhere plausible and the reconciliation closes. The package looks clean. Six months later somebody discovers a suspense balance with fourteen items in it and no memory of any of them.
The alternative is straightforward: unresolved items are escalated, in writing, with whatever documentation exists, in the period they arise. It makes the package look less tidy and makes the file dramatically more sound.
Prioritization matters as much as completeness. A list of thirty questions with no ordering is a list the reviewer will not get through. Three things that need a decision, followed by twenty-seven that need acknowledgement, is a list that gets actioned.
The close checklist
What was done, by whom, and when — including the review step.
This is the part that turns a set of documents into evidence of a process. It should show that reconciliations were performed, that recurring entries were posted, that the agings were tied out, and that somebody other than the preparer reviewed the file before it was released.
The checklist is also the thing that makes the work survive turnover. When staffing changes, the next person inherits a documented procedure rather than a guess.
What does not belong
- Every report the software can generate. Volume is not thoroughness.
- Screenshots without context. If it is worth including, it is worth labelling.
- Anything the reviewer would have to reconstruct to trust. Either do the work or flag that it was not done.
The test
A well-built close package can be reviewed in a fraction of the time it took to build. If your review takes nearly as long as preparing it would have, something in the package is wrong — usually the open items are missing, the schedules are absent, or the reconciliations are asserted rather than evidenced.
That ratio is the only meaningful quality measure. It is also the one worth asking any prospective bookkeeping provider about directly: how long should reviewing your close package take me? A provider who has thought about the question will have a specific answer.
Every file apar closes goes through a documented review under CPA supervision before it reaches your firm. See what is in our close package.