Those who own polling companies use them as a tool of influence and propaganda, nothing more.
538's polling averages on election day put Biden at 51.8% of the vote and Trump at 43.4%. The margins of error were roughly 3 points either way.
Biden received 51.3% of the popular vote and Trump 46.9%.
https://projects.fivethirtyeight.com/polls/president-general...
In any case, even if the polls were systematically off, and they may have been, that doesn't justify calling them all propaganda, or bs. It isn't even evidence of it. Because, as it happens, polling people is effing hard.
Look, there clearly are some "bullshit polls", and there are partisan polls paid for by candidates which tend to be favorable to them, and there a good polls with consistent bias, a kind of "house" effect as a consequence of their sampling methodology, etc.
That's why Nate Silver does averages of polls, btw, and weights polls in the average by their accuracy in previous elections.