Challenge 1: A Two-Level Sort — Solution Walkthrough

1. Select the full range A1:C20 (including headers).
2. Open Data > Sort.
3. Make sure "My data has headers" is checked, so the dialog shows
   column names instead of column letters.
4. First sort level: Column = Department, Order = A to Z.
5. Click "Add Level."
6. Second sort level: Column = SalesTotal, Order = Largest to
   Smallest.
7. Click OK.

Expected result:

All rows are grouped together by Department, with the departments
themselves appearing in alphabetical order. WITHIN each department
group, the rows are further ordered so the highest SalesTotal appears
first and the lowest last.

Explanation:

The Sort dialog applies levels in the exact order they're listed —
the first level is the PRIMARY sort key, and every level after it
only breaks ties within groups the level above it already created.
Here, Department is decided first (grouping every row into
alphabetical department blocks), and SalesTotal is only used to
decide ordering INSIDE each of those already-formed blocks — it never
mixes rows from different departments together by sales figures.

Notes:
- Selecting the FULL range (A1:C20), not just one column, is what
  keeps each employee's Name, Department, and SalesTotal moving
  together as one row — exactly the mistake this chapter's own
  warning box describes if only a single column were selected and
  sorted in isolation.
- Swapping the order of the two sort levels (SalesTotal first,
  Department second) would produce a completely different, far less
  useful result — a single list ranked by sales company-wide, with
  department membership scattered throughout rather than grouped.