A social media website decided to change its name and wants to see how users feel about the new name. To get an accurate read on how users of the website feel, they put a poll on their website. In order to combat the idea of 'convenience sampling,' this poll is only issued to some registered users at random, rather than all users. That way they are doing a proper 'random sampling' rather than biased convenience sampling. Despite this, the poll is accused of being biased. What is one way this poll might still be biased? (You may suggest potential situations I didn't bring up.)