Best on desktop
The spreadsheet workspace needs more room than this screen offers. Switch to a laptop for the full experience.
You're a marketing analyst pulling tomorrow's territory readout. The order log at A2:B25 records (Region, Customer) for every order — and customers typically order multiple times per region across the quarter, so the table has heavy duplication. You need the count of distinct customers for each of the three regions in D2:D4.
Two acceptable approaches — pick whichever you're more comfortable with under the timer:
=COUNTA(UNIQUE(FILTER($B$2:$B$25, $A$2:$A$25=D2))) — three function calls.=SUMPRODUCT(IFERROR(($A$2:$A$25=D2) / COUNTIFS($A$2:$A$25, D2, $B$2:$B$25, $B$2:$B$25), 0)) — for each row in the target region, contribute one over the count of (target region, that customer) combos. Each unique customer sums to exactly 1.The IFERROR is necessary in the legacy version because COUNTIFS returns 0 for rows whose customer doesn't appear in the target region — division by zero would otherwise produce #DIV/0! and corrupt SUMPRODUCT.
Your task:
Graded cells: E2, E3, E4.