Data Preparation and Blending
Data Preparation and Blending questions test whether you understand how Alteryx changes data, not just which tool icon looks right. Core concepts: field types (numeric sizes, fixed decimal, string vs. V_String), the difference between the Select tool (metadata changes: rename, type, order, deselect) and Auto Field (infers types, can silently change numeric to string), and the Data Cleansing tool (whitespace, nulls, case, unwanted characters). Know the Join family: Join (inner by default, with L/R/unjoined outputs), Join Multiple (requires a common key or cross-join), Union (by name vs. by position - a classic trap), Append Fields (cartesian), and Find Replace (append vs. replace, match case). Blending across sources usually involves Join on cleaned keys or Fuzzy Match for imperfect strings; be ready to explain why a join drops rows (type mismatch, trailing spaces, null keys).
Traps to watch: Union by position mismatches columns when field order differs; Join treats a numeric 100 and string "100" as non-matching; Auto Field converts leading-zero ZIP codes; the Multi-Row Formula requires grouping and sort order to behave; Sample is random unless you set a seed. Also remember Select can change a field to a smaller type and truncate data silently.
Rule of thumb: Inspect with Browse before and after every join. Worked example: two files, one stores CustomerID as Int32, the other as String. An inner Join returns zero matched rows. Fix: add a Select on the string side to convert to Int32 (after trimming). If trimming is needed, use Data Cleansing first - conversion fails on spaces.