Parity Reconciler
Compare two CSV exports, locate the gaps, and see what to check next. Patterns are clues, not confirmed causes.
1 · Drop two exports
Export daily CSV/TSV data with a header row, date, channel and one metric. Use the same metric definition and date range on both sides. Export detail rows without totals; duplicate rows are not automatically removed. Match account currency and timezone in the source platforms before exporting.
Download a synthetic CSV example · Read the practical guide (English)
No data handy? Use the sample button in the workbench header.
2 · Map columns & choose what to compare
Source A columns
Source B columns
Confirm the metric column on each side and its meaning below. “Results” or “conversions” does not necessarily mean installs. Matching numbers do not establish matching definitions.
What we found
This tool locates observed gaps and suggests checks. It cannot confirm the cause or determine which source is correct. Missing settings remain unknown.
Daily breakdown
Rows that deviate most from the overall gap
These rows are statistical outliers vs. the overall gap — start any manual digging here.