Affiliate Reporting: Check Decimal Separators Before Summing Revenue

A revenue column can look consistent while its imported values mean something different from the source. Performance marketing media buying teams combining affiliate exports should therefore check number formatting before comparing campaign totals. A separator mismatch can change the scale of a value or leave it as text that a calculation ignores.

The method below is a proposed import-review workflow. It concerns how numeric text becomes a number; currency conversion, commission eligibility and payment settlement require separate checks.

Establish the source convention

Record the export's documented numeric format and inspect representative raw values before a spreadsheet or reporting tool transforms them. Under a decimal-point and comma-grouping convention, "1,234.56" represents one thousand two hundred thirty-four and fifty-six hundredths. Under a decimal-comma and period-grouping convention, the same amount can be written "1.234,56".

A short string such as "1,234" is ambiguous without the source convention: it could represent a grouped integer or a decimal value. Do not decide based only on the reader's location, the currency symbol or the largest value in the column. If documentation and export settings do not settle the ambiguity, flag the rows for clarification.

Preserve raw text alongside parsed values

Keep the original file unchanged. In a working copy, retain a raw-value field, a parsed-number field and a parsing-status field. This makes it possible to review a transformation without reconstructing the original text from a rounded display.

• Document the expected decimal and grouping symbols for each source.
• Handle spaces, signs and currency labels only according to that documented format.
• Flag unexpected patterns instead of silently deleting punctuation.
• Keep empty, invalid and valid zero values distinct.
• Record the import setting or transformation used.

Removing all commas is not a universal cleanup rule. It would convert a decimal-comma value such as "12,50" into "1250" rather than 12.50. A consistent-looking result can therefore be consistently wrong.

Check known values and control totals

Use a small test set with independently known meanings before importing the full report. In a hypothetical decimal-comma export, "12,50" and "7,25" should produce 12.50 and 7.25, with a sum of 19.75. This example checks interpretation; it is not campaign performance data.

Count source rows, successfully parsed rows and exceptions. Then reconcile against a source total when one is available with the same filters, time range, currency and transaction status. A matching total is helpful but does not replace individual checks: two errors could offset each other.

Make exceptions visible in campaign decisions

If unresolved rows remain, disclose their count and exclude the affected comparison from a scaling decision until the ambiguity is resolved. Do not relabel a partial total as complete revenue. Keep the source convention with the dataset so that the next export does not depend on someone's memory.

For affiliate media buyers comparing ecommerce CPS and virtual-product campaigns, a reproducible numeric import is a foundation for analysis. It lets the team discuss actual campaign differences instead of an unnoticed change in punctuation.