=IF(A2="", "", COUNTIFS(A$2:A2, A2, B$2:B2, "<>"))

=IF(ISBLANK(A2),"",COUNTA(A$2:A2))

This counts how many times the current category value has appeared so far in the expanding range. When the category changes (e.g., from “Fruit” to “Vegetables”), the count resets to 1. This creates perfect nested numbering: Fruit: 1, 2, 3; Vegetables: 1, 2; Dairy: 1.

Using LET (Excel 365):

This is a form of window function (similar to ROW_NUMBER() OVER (PARTITION BY Category) in SQL). It demonstrates that Excel’s grid can perform relational database operations without a database engine. This technique is invaluable for creating outlines, bill of materials (BOM) exploded views, or numbered lists inside pivot table source data. 4. The Advanced Synthesis: Combining Visibility and Hierarchy The ultimate challenge: number visible rows only, restarting the count per group, after a filter. This requires an array formula (or the new LET and FILTER functions in modern Excel).

that also ignores blanks:

At first glance, numbering cells in Excel appears trivial. The user reaches for the fill handle, drags down, and Excel autocompletes a sequence (1, 2, 3...). However, this primitive method shatters the moment the data structure becomes irregular. What happens when rows are empty? What if you need to count only visible rows after a filter? What if the numbering must restart based on a change in a category?

This is where becomes essential. It transforms Excel from a static grid into a dynamic database engine. Conditional numbering is not about counting cells; it is about assigning an incremental identity based on logical tests. This essay explores the three primary paradigms for conditional numbering in Excel: the COUNTIF expanding range, the SUBTOTAL function for filtered data, and the COUNTIFS multi-condition ranking. 1. The Classic Sequential Condition: The Expanding Range The most fundamental conditional numbering problem is: "Number only the rows where Column A is not empty, ignoring blanks."