7/ 9
POWER QUERY CONTROL VALIDATION
Verify your Power Query reconciliation before you trust the refresh.
Check duplicate and blank match keys, unmatched records, invalid mapped amounts, and amount differences in an existing recurring reconciliation. Review source row counts and valid amount totals alongside the exceptions.
For reconciliations you already maintain. ReconVerify validates selected exports or pasted data; it does not build, refresh, or automate the Power Query reconciliation itself.
Validation run · Aug 16, 2026
Needs review Reconciliation check
2,479
one-to-one keys with valid amounts $0.00
Exact matching Issues to review
2 controls need attentionDuplicate match keysSource B
2 match keys appear more than once in Source B.
Amount mismatchBeyond tolerance
One unique matched key has different mapped amounts in Source A and Source B.
125.00→120.007 controls passedThe remaining key, amount-format, and unmatched-record controls passed.
Existing Power Query reconciliations Deterministic key and amount controls Human review stays in the loop Files stay in your browser
CURRENT CONTROL SET
A refresh can succeed while the reconciliation quietly changes.
ReconVerify currently runs deterministic checks against the mapped match keys and amounts. It also reports source row counts and valid amount totals so you can review the input populations alongside the exceptions.
Key completeness
Blank values in the mapped match-key column.
Cardinality
Match keys that occur more than once on either side.
Coverage
Keys present in Source A but not Source B, and vice versa.
Amount validity
Mapped amount values that cannot be parsed by the current numeric rules.
Amount matching
Differences on unique matched keys beyond the tolerance you set.
VERIFICATION GUIDE
What should you verify after a Power Query reconciliation refresh?
A successful refresh is not the same thing as evidence that a recurring reconciliation behaved the same way as the prior run. Verification can include row counts, match cardinality, coverage, amounts, schema, data types, and retained transformation errors.
Microsoft documents that Power Query merges depend on selected join keys and join kind, and that schema or data-type changes can produce errors or unintended results. Merge queries Data-source errors
What is a cardinality failure?
A relationship expected to be one-to-one becomes one-to-many or many-to-many. Duplicate join keys can multiply output rows when merged matches are expanded.
What is a completeness failure?
A transaction expected to participate in the reconciliation is missing or remains unmatched. Review unmatched records on both sides instead of relying only on a final net difference.
What is an amount-integrity failure?
Source totals, matched amounts, or output totals differ from the expected relationship. Row-level differences and signed totals show whether the result was reached for the expected reasons.
What is schema drift?
A required column is missing, renamed, or unexpectedly added. Structural changes should be explicit review conditions rather than invisible input changes.
What is data-type drift?
A field expected to be a number, date, text value, or another defined type changes type or retains conversion errors. Required types and parse errors deserve their own controls.
SYNTHETIC EXAMPLE
A duplicate key can change the result without changing the query logic.
Assume transaction IDs are expected to be unique on both sides. If A03 appears once in the ledger and twice in the bank data, expanding the merge can create two output rows for the ledger transaction.
- Expected relationship
- 1 → 1
- Observed for A03
- 1 → 2
- Review condition
- Cardinality failure
BEFORE YOU TRUST THE REFRESH
Retain evidence for the assumptions that matter.
A reviewer should be able to see what was checked, which exceptions remain, and what changed from the expected recurring behavior.
- Source assertionsRow counts and signed totals for each input.
- Matching behaviorDuplicate keys, one-to-many matches, and unmatched records.
- Output assertionsRow counts and amounts compared with the expected relationship.
- StructureRequired columns, expected schema, and required data types.
- Transformation errorsRetained errors that need diagnosis rather than silent removal.
- Change reviewFailed controls, reviewed exceptions, versions, and reviewer disposition.
WORKFLOW
Keep the reconciliation. Add a verification layer.
Refresh
Run the Power Query reconciliation you already maintain.
Verify
Load the two resulting datasets and run deterministic controls on keys, coverage, and mapped amounts.
Review
Inspect exceptions, row-count and amount-total metrics, and keep the final judgment with the responsible reviewer.
DESIGNED FOR FINANCE OPERATIONS
Control evidence without pretending software replaces judgment.
ReconVerify is for recurring accounting and finance-operations workflows where an existing Excel Power Query reconciliation matters enough that silent changes deserve explicit controls.
- Typical use
- Month-end, bank and ledger, recurring operational reconciliation
- Decision model
- Deterministic checks, exception review, human sign-off
- Data handling
- Supported inputs are read locally in your browser, with no file upload action
BROWSER-LOCAL VALIDATOR
Try the product without creating an account or uploading your files.
Start with Excel workbooks, CSV, TSV, delimited text, or data pasted from Excel. ReconVerify reads supported inputs locally in your browser, lets you map the reconciliation fields, runs the current key and amount controls, and produces a local validation report you can view or download.
Try ReconVerifyCurrent browser beta checks blank and duplicate match keys, unmatched keys, invalid amounts, and amount differences beyond your chosen tolerance. Row counts and valid amount totals are reported as review metrics.