Is the content being mailed the same? are you mailing the same people again before and after alex was created? are you split-testing -- 1/2 recipients get the email without alex in CC? You might need some more experiments to attribute it to pure CC.