Trial balance agreement
Total debits and credits agree with the source, and net to zero, for every period in scope. This is the top-level check and the least interesting one.
Platform · trust
Ask any vendor how you will know a migration was complete and correct. The good answer is a per-period, per-account variance report that ties the loaded balances to the source system and that you keep. Most projects produce a sign-off email instead, which proves somebody was satisfied rather than that the numbers agree.
184 consecutive days tied · variance $0.00
What it does
Total debits and credits agree with the source, and net to zero, for every period in scope. This is the top-level check and the least interesting one.
Every account balance matched period by period, so a difference is located rather than known to exist somewhere in the aggregate.
AR and AP detail ties to its control account on both sides. An aggregate that agrees while the detail does not is the failure mode this catches.
Any difference resolves to the specific source transaction causing it, with a reason — unmapped account, timing difference, classification disagreement, or genuine error.
Each period in the migration scope, not a spot check on the most recent one. A conversion that ties in December and not in March is not converted.
Yours in an open format, retained independently of us. It is the document you hand an auditor who asks how you know the conversion was complete.
The first runs of a reconciliation proof almost always show differences, and customers occasionally read that as a bad sign. It is the opposite. Timing differences, unmapped accounts, and classification disagreements are exactly what a migration has to surface, and finding them in week one costs a fortnight of adjustment.
The same differences discovered at a cutover weekend cost the cutover. Every one of them exists whether or not anybody looks; the only variable is when.
A reconciliation that produces a report somebody reviews under deadline pressure is a formality. Ours gates: the migration does not advance past a period that does not agree, and a variance sends the work back to mapping rather than being waived.
That is enforced on our side as much as yours — we will not sell a ledger cutover until three consecutive closed months have tied at zero variance. It applies to our revenue, which is the only version of a gate that means anything.
It is not an audit and it is not an opinion on whether your source data was right. If your QuickBooks file contains an error, a correct migration reproduces that error faithfully and the proof will show agreement. Agreement means the conversion was faithful, not that the underlying books were correct.
Where we notice something that looks wrong in the source — an accumulated depreciation balance that does not tie to the register, a suspense account with years of activity — we raise it separately as a finding rather than folding it into the migration.
If a shadow ledger has been running against your books for months, the proof already exists as a historical record rather than being produced for the cutover. That is the strongest version of this: months of daily agreement rather than a reconciliation performed once, under pressure, by the party being paid for the conversion.
Limits
Faithful conversion of an incorrect balance produces agreement. The proof shows the migration was accurate, not that the original books were right.
It is evidence your auditor can use and not a substitute for their procedures. We produce the artefact; they form the view.
Standard scope is opening balances plus one to two years. Where you migrate more, the proof covers what was migrated and we state the boundary explicitly.
Questions
A per-period, per-account variance report you keep. We will send an anonymised sample so you know what to ask for.