There may be times when you need to get the top 10 (or n) rows from each group of your dataset to satisfy a business requirement. There are a few ways this can be done, but none so elegant as when you use RANK().

RANK() effectively applies a ranking to each rows in your dataset, based on a specified grouping and order. This can then be used to select the Top n rows by using an appropriate where clause. Here’s an example:

   1: USE Adventureworks

   2: go

   3:

   4: WITH OrdersCTE

   5: AS

   6:

   7: (

   8: Select

   9:     salespersonid,

  10:     customerid,

  11:     SUM(subtotal) Total,

  12:     RANK() OVER(PARTITION BY salespersonid ORDER BY SUM(subtotal) desc) Ranking

  13: from Sales.SalesOrderHeader

  14: Where

  15:     salespersonid is not null

  16: Group by

  17:     customerid,

  18:     salespersonid

  19: )

  20:

  21: Select

  22:     salespersonid,

  23:     customerid,

  24:     Total,

  25:     Ranking

  26: From OrdersCTE

  27: Where

  28:     Ranking <= 10

In the above example, I needed to find the top 10 Customers for each Sales Person based on the sum of all order subtotals (using AdventureWorks 2005). Here are the results:

Ranking Results

You can read more about the RANK() function here, and for more ranking functions see DENSE_RANK(), NTILE() & ROWNUMBER().