Research question: when a bookkeeping VA deletes a QuickBooks Online journal entry, does the audit log retain a usable record of the entry, and which report history still shows its effect?

Testing Whether Deleting a QuickBooks Online Journal Entry Removes It from the Audit Log and Report History

A bounded test method for a bookkeeping VA checking what remains in QuickBooks Online after a journal entry is deleted, using baseline reports, audit log exports, and evidence limits.

Testing Whether Deleting a QuickBooks Online Journal Entry Removes It from the Audit Log and Report History research thumbnail
On this page Jump to section

Question and scope

When a journal entry is deleted, the ledger stops showing it. The open question is what a reviewer can still prove afterward: whether the audit log records the deletion with enough detail to reconstruct the entry, and whether previously saved or exported reports preserve the amount that the ledger no longer carries. A bookkeeping VA meets this question after a duplicate entry is removed, during a cleanup, or when an accountant asks why a balance changed between two closes. The answer decides whether deletion is an acceptable correction or whether the entry should have been reversed.

The scope is one QuickBooks Online company file, one journal entry, and one user role with permission to delete. It covers entries deleted from the journal entry list and entries removed through a supporting workflow, but it excludes entries voided rather than deleted and entries changed by a period close. The test compares the ledger, the audit log export, and reports generated before and after the deletion. It does not attempt to recover the entry from a backup.

Why deletion differs from reversal

A reversal adds an offsetting entry, so both the original and the correction remain visible in the ledger. Deletion removes the original, which makes the account balances look as if the entry never existed. That difference matters when a reviewer needs to explain a change to a prior period, because a reversal carries its own date, memo, and approval while a deletion may leave only a system event. The test has to show which of those states the file actually presents.

The distinction also affects report history. A report run after the deletion reflects the current ledger, while a report exported before the deletion reflects the earlier state. If nobody saved the earlier export, the two states may not be comparable. The test therefore treats the baseline export as the control, not as an optional convenience. Without it, a reviewer can see that a balance moved but cannot show the amount that was removed.

Population and baseline

Build the population from journal entries deleted during a defined window, drawn from the audit log when the log can be filtered by user and date. Record the entry number, posting date, amount, accounts affected, preparer, and the date and time of deletion. Include at least one balanced entry, one entry with multiple lines, and one entry that touched a bank or tax account. Exclude entries removed as part of a company file conversion, because the conversion may rewrite identifiers.

Freeze the baseline before deleting anything. Export the trial balance, the general ledger detail for the affected accounts, and the profit and loss and balance sheet for the periods on either side of the entry date. Save the audit log export for the same window, along with the report run time and the user who ran it. Capture the entry detail itself, including every line, class, location, and memo, so the reconstruction test has a known target.

Test setup

Work in a copy of the company file when the subscription and the client permit it. A deletion in the live file changes balances that may already be reconciled or reported, and recreating the entry may not restore the original identifier. If a copy is not possible, schedule the test outside the close window, obtain written approval, and record every setting before the change. The test must be reversible without guessing which values were original.

Change one variable at a time. First delete a simple balanced entry and compare the ledger and audit log. Then repeat with a multi-line entry and with an entry that touches a bank account. Finally test an entry that was created and deleted by a different user role. A test that changes the entry type and the user at the same time cannot show which factor changed the evidence.

Procedure

Run the baseline comparison before deleting the entry. Reopen the saved reports and confirm that they still match the exports, which catches ordinary drift from later activity. Record the current balances for every affected account and the audit log rows that mention the entry. Then delete the entry through the normal interface and save the confirmation message, if the product shows one.

Immediately rerun the trial balance, general ledger detail, and financial statements with the same date range and accounting method. Compare the affected account balances with the baseline and record the exact difference. Export the audit log again and search for the entry number, the amount, the posting date, the preparer, and the word delete. Note which fields remain and which are missing, rather than summarizing the result as present or absent.

Repeat the search in a saved report and in a report that was never exported before the deletion. The first shows whether the earlier export preserves the entry, and the second shows whether the current report can still produce it. Then recreate the entry from the captured detail and compare the new identifier with the old one. This step shows whether the file can restore the accounting effect, even when the evidence trail cannot restore the original record.

Definitions and measures

A retained audit record means the audit log still shows the deletion event with enough fields to identify the entry, its amount, its date, and the user who removed it. A partial record means some fields remain but the entry detail cannot be reconstructed from the log alone. A missing record means the log no longer shows the event, or shows only a generic activity line that does not identify the transaction. Record the state for each field so the memo does not collapse a partial result into a pass or a failure.

Measure the result at three levels. The ledger level shows whether the balances changed by the expected amount. The audit level shows which fields survived and whether the event can be tied to a named user. The report level shows whether an earlier export preserves the entry and whether a current report can reproduce it. Report all three, because a clean ledger can coexist with a weak evidence trail.

Worked example

Suppose a client has a 2,400 dollar journal entry that moves insurance expense from the wrong department to the right one. The entry was posted on the last day of the month and included two lines, one debit and one credit, with a class on each line. A VA deletes it after the client says the allocation was wrong, then recreates it with the corrected class. The ledger ends the month with the same total expense and a different class split.

The test compares the baseline trial balance with the post-deletion trial balance and confirms that the class totals moved as expected. It then searches the audit log for the original entry number and the amount. If the log shows the deletion but not the line detail, the reviewer can prove that something was removed but cannot rebuild the original allocation without the saved export. That gap belongs in the close note, along with the replacement entry number and the client's approval.

Evidence standard

The evidence set includes the baseline reports, the original entry detail, the audit log export before and after the deletion, the deletion confirmation, and the replacement entry if one was created. Each claim in the memo should point to a specific artifact with a timestamp and the account or entry it affects. A statement that the audit log retains the deletion is credible only when the export shows the event and the identifier. Keep the raw exports rather than only a summary table.

Add at least two deletion examples so the conclusion rests on more than one entry. Include one entry that touched a bank account, because those entries often carry reconciliation implications that a simple expense reclass does not. If the report cannot be reproduced because the period is locked or the log export is incomplete, say so and lower the confidence of the conclusion. Do not fill the gap with an assumption about how the product should behave.

Interpretation limits

A result in one company file does not prove that every QuickBooks Online edition or user role behaves the same way. The outcome can depend on the subscription level, the permissions attached to the user, the audit log retention window, and whether the entry was deleted before or after a period close. Record the product edition, the user role, and the test date so a later reviewer can judge whether the result still applies. A second file with a different role is a useful follow-up test.

The test also cannot decide whether deletion was the right correction. That question depends on the client's accounting policy, the materiality of the entry, the status of the period, and the accountant's view on prior-period adjustments. The test can show what the file retained and what the reports can still prove. It cannot substitute for a correction policy, and the memo should state that boundary.

Review cadence

Repeat the test after a permission change, after a period close, and before a cleanup that is expected to delete entries. A monthly spot check is reasonable for a client with several preparers, while a client with one preparer and a stable file can review quarterly. Put the next review date in the client's close calendar rather than in a personal reminder. Record the previous result beside the new one so a change in retention behavior is visible immediately.

Close the review with a named owner, the exact report settings, and a date for the next check. Link the result to the journal entry reversal guide and the journal entry support log so the next reviewer starts from the same evidence. If deletion left an unusable trail, convert the finding into a written correction policy with an approval step. The test explains the state; the accountant decides what belongs in the close.

Sources