SQL Query to Find Most Popular Instances of Data

Question

I am trying to build a query that will sort occurrences of the field DESCR254. So for example, test1 is used by 3 different USERIDs and test2 would be the second most. Does anyone know how I would construct this query given there are many more rows of data and many different instances of USERID and DESCR254?

select * from sysadm.PS_IS_STATS_URLS

PS_IS_STATS_URLS

USERID   DESCR254 
ZX       test1
ZX       test2
ZC       test1 
ZB       test3
ZA       test1
ZA       test2

I would want the result to look something like this:

DESCR254    COUNT
test1       3
test2       2
test3       1 

Answer 1

SELECT DESCR254, COUNT(*) AS COUNT
FROM sysadm.PS_IS_STATS_URLS
GROUP BY DESCR254
ORDER BY COUNT(*)

Answer 2

To count only accesses from different userids as your question suggest, you need to use COUNT DISTINCT;

SELECT DESCR254, COUNT(DISTINCT USERID) "COUNT"
FROM PS_IS_STATS_URL
GROUP BY DESCR254
ORDER BY "COUNT" DESC;

An SQLfiddle with a non counted duplicate to test with.