Merge appends rows with the same schema. Join combines columns from records with matching keys. Choosing the correct operation prevents accidental row loss or multiplication.
Merge more records with the same columns
Both headers must match exactly and be in the same order. The tool preserves duplicates and appends the second table’s rows. If an export has an extra column or different spelling, align the schema first instead of assuming equally positioned fields mean the same thing.
Join related records by an explicit key
Key matches are exact parsed strings, so 001 differs from 1. A left join retains unmatched left records with blank right fields. An inner join keeps matches only. Output columns receive left/right prefixes so similarly named columns remain distinguishable.
Count repeated keys before summing
A key occurring twice on the left and three times on the right creates six joined pairs. That is expected relational behavior, not necessarily a parser error. Inspect row counts, unmatched records and whether the intended key should be unique before summing any joined amount. The browser limits join output to 10,000 rows.
Worked example
Scroll the table horizontally on small screens to see every column.
| Left occurrences | Right occurrences | Joined pairs |
|---|---|---|
| 1 | 1 | 1 |
| 2 | 3 | 6 |
A join can legitimately contain more rows than either source table.
Sources and checks
Examples are illustrative and can be checked with the linked tools. Read each tool’s assumptions before applying its result. How we verify methods.