Optimize — Replace 16 SUMIFS Cells with One Spill

You're a marketing analyst optimizing a 4×4 cohort retention matrix. The current approach is 16 separate cells, each computing returns / cohort_size for one (cohort, month) combination. Adding a 5th cohort means 4 more cells; adding a 5th retention month means 4 more cells; both means 9 more cells.

The pre-aggregated source table:

• A2:A5 — cohort label (Jan, Feb, Mar, Apr)
• B2:E5 — return counts at months M+1 / M+2 / M+3 / M+4 (4×4 block)
• F2:F5 — cohort size (the cohort's first-month customer count)

Output area is empty: H1:K1 already has the headers (Retention_M1 → Retention_M4).

Your task:

1. In H2, write a single 2D array operation that divides the 4×4 returns block by the 4×1 cohort-size column. Modern Excel broadcasts the column division across each row of the block; the result spills into H2:K5.

The pre-365 alternative is 16 =B2/$F$2 cells, four times across, four times down. The spill version is one expression that scales linearly: 5×6 cohort matrix is the same one cell.

Graded cells: H2 (Jan × M+1 → 0.5), J3 (Feb × M+3 → 0.3), K5 (Apr × M+4 → 0.1).