Let's say you have two or more identical strings of data. In this example, the text "Hello" is found more than once.

Hello John Hello Amy Hello Sal

A simple sed statement will replace every occurrence of the world "Hello". 

~]# echo "Hello John Hello Amy Hello Sal" | sed "s|Hello|Goodbye|"
Goodbye John Goodbye Amy Goodbye Sal

The following sed statement can be used to only replace the last occurrence of "Hello".

~]# echo "Hello John Hello Amy Hello Sal" | sed "s|\(.*\)Hello|\1Goodbye|"
Hello John Hello Amy Goodbye Sal