VSThiran

Data & Spreadsheets

Reconcile Two Sets of Records

Free to useNo sign-up requiredNo watermarkRuns in your browser

Two systems that should agree and do not: a ledger against a bank export, a supplier statement against your purchase records, an invoice list against what was actually paid.

Point this at both, tell it which column is the reference and which is the amount, and it reports what matched, what is missing from each side, and every amount difference larger than the tolerance you set - with the totals from both sides and the gap between them at the top.

How this tool works

  1. Add both datasets

    Two XLSX, XLS or CSV files.

  2. Say which columns mean what

    The reference on each side, and optionally the amount and a description. The two files do not have to use the same headings.

  3. Set the tolerance

    How large an amount difference to ignore. Rounding noise is not a finding; anything larger is reported with its own figure.

  4. Reconcile

    The summary shows both totals and the difference, with every exception listed below.

  5. Download the workbook

    A sheet for each outcome, each carrying how the reference was matched.

How it works

References are matched exactly first, then again after tidying up spacing and capitalisation, and only then - if you turn it on - by similarity. A close match can never displace an exact one.

An amount difference outranks a description difference. If the money is wrong, that is the finding, whatever else also differs on the row.

Tolerance applies only to the amount columns and is an absolute figure, so a tolerance of 0.01 accepts a one-penny rounding difference and reports everything larger. Every difference is reported with its own value rather than a pass or fail.

Every row of both datasets appears exactly once in the result, so the counts always add up. A reference that repeats within one dataset is reported as a duplicate rather than being reconciled twice.

The whole comparison runs in a background thread in your browser, so the page stays usable and neither file is uploaded.

What each outcome means

MatchedFound on both sides, and the amount agrees within your tolerance
Amount DifferenceFound on both sides, but the amounts differ by more than the tolerance
Field DifferenceAmounts agree, but a description column differs
Missing in APresent in the second dataset only
Missing in BPresent in the first dataset only
DuplicateThe reference appears more than once, so it cannot be paired reliably
Needs ReviewA blank reference, or two candidates too close to choose between

Frequently asked questions

Is this reconciliation tool free?

Yes, free with no sign-up.

Are my files uploaded anywhere?

No. Your files stay on your device. Both are read and reconciled by your browser, which matters when the data is financial.

Do both files need the same column names?

No. You choose the reference, amount and description column on each side independently.

What tolerance should I use?

Enough to absorb rounding, and no more. 0.01 is a sensible default for currency. Set it to zero to see every difference however small.

What happens if a reference appears twice?

It is reported as a duplicate rather than reconciled, because there is no way to know which row on one side belongs to which on the other.

Can it reconcile records with no shared reference number?

Turn on close matching and it will match on similar references. That is a fallback, not an improvement on a real shared identifier, and anything too close to call is reported for review rather than reconciled on a guess.