Data Blending and Joining
Combining streams
The Join tool takes two inputs and connects them on one or more key fields, producing three outputs: Join (matched records, left and right fields side by side), Left (unmatched left records), and Right (unmatched right records). It behaves as an inner join on the J output; keep the L and R anchors when you need to audit what failed to match. When both inputs contain a field with the same name that is not the join key, Alteryx renames the right-side field with a Right_ prefix so both survive.
Other blending tools: Join Multiple connects three or more streams and can output unjoined records; Union stacks streams, either by name (fields matched by name, mismatches become nulls) or by position (fields matched in order); Append Fields adds every right-side field to every left row with no key at all; Find Replace looks up values and can replace or append; Fuzzy Match scores near-matches using a match style and produces a MatchScore; Make Group assigns group IDs to those pairs.
Traps
Join keys must be the same data type - joining a string to a numeric field silently yields no matches or an error. Trailing spaces and case differences break joins, so cleanse keys first. Union by name silently produces nulls when names differ, and Union by position fails when the column order differs. Join by Record Position is a trap in disguise; it is order-dependent, not key-based.
Rule of thumb
Before joining, verify key type, trim whitespace, and normalize case. After joining, check the L and R counts - a non-empty Right output usually means unclean keys.