Practice — Mild · The Stack and the Lookup

Skills focus. Append · Left Outer merge · selective expansion · verifying row counts at each stage.

Ungraded. Every checkpoint is a verified figure.

Tasks

Import all six sales_export_2026-*.csv files (Get data → Text/CSV, six times — or try the Folder connector on the directory if you are feeling professional).
Append Queries as New, selecting all six. Name the result Sales H1. Checkpoint: 26,139 rows.
Import products.csv. Before merging, profile its ProductID column. Checkpoint: 240 rows, 240 distinct — a clean key. This ten-second check is what guarantees the merge cannot multiply rows.
Merge products into Sales H1 on ProductID, Left Outer. Expand Category only. Checkpoint: still 26,139 rows.
Close & Apply. Build a bar chart of Quantity by Category. Checkpoint: Outdoor leads at 9,439 units; Decor trails at 6,191.

Hint — my appended table has extra half-empty columns

Column names must match across files for append to align them. Check each monthly query's headers — if you renamed one while exploring, the append kept both versions. Delete the rename step and append again.

Reflection

Step 3 took ten seconds. What merge disaster does it rule out, and what would that disaster have looked like in your row count?
Why Left Outer here, when every ProductID matches and Inner would give the same 26,139?