Every character is first assigned a universal number called a code point. UTF-8 then converts that number into a sequence of one to four bytes so it can be stored or sent as raw data.