Level 5 · Returns by reason and month, biggest first

Dates are too fine-grained for a summary: 28 returns, each on its own date over three months, tell you little until they are rolled up. Analysts group dates into months (or weeks, quarters, years) and then sort the result so the biggest item comes first.

Grouping dates: two ways

In Excel, once a date field is in a PivotTable, you can right-click one of its dates and choose Group, then tick Months (and Years, so May 2024 and May 2025 stay apart). Excel then builds the month labels itself.

The other way works everywhere, including Google Sheets and this practice sheet: add a helper column to the data and let the PivotTable use it like any other field.

F1   month
F2   =TEXT(C2,"yyyy-mm")

Filled down to row 29, every return gets a label like 2025-05. The yyyy-mm format sorts correctly as text, which May 2025 would not. Because F1 has a header and sits right next to the data, the PivotTable’s source grows to include it, exactly as Excel’s would when the data is formatted as a Table.

Count and sort

With month available, build: Rows reason, Columns month, Values return_id as a Count. Then set Sort rows to Largest to smallest. In Excel that is a right-click on a value in the Grand Total column → Sort → Sort Largest to Smallest.

Count of return_id   month
reason          2025-04   2025-05   2025-06   Grand Total
damaged               4         5         8            17
changed_mind                    6         3             9
late                  1                                 1
wrong_item            1                                 1
Grand Total           6        11        11            28

The sort uses each row’s Grand Total. Ties (late and wrong_item, 1 each) keep A→Z order.

Reading it

Damaged items are 17 of the 28 returns and the only reason that grows every month: 4, 5, then 8 in June. In refunds the gap is bigger still: damaged returns cost 11,194.91 of the quarter’s 16,760.28 (switch Values to Sum of refund_amount to see it). That points at packaging or the carrier, and it is worth fixing before July, Outdoor’s busiest month.

Common mistakes

• Dates straight into Columns. Without grouping you get one column per day.
• A month label that sorts wrong. TEXT(C2,"mmm") gives Apr, Jun, May in A→Z order. Use yyyy-mm, or month numbers.
• A helper column without a header, or with a gap before it: the PivotTable can’t use it.
• Sorting by the labels when the question is “biggest first”. A→Z is the default; change it.