*clears her throat and looks around lecture theater*
Alright… okay, serious.
It is not really clear, what the Twitter data means. It seems so far, that there is a correlation between Twitter chatter and people tuning in, meaning lot of buzz on Twitter makes people curious and tune into a program – like when people on the street run and look in one direction we can’t help it but follow, curious to see what is going on. It is assumed that as well Twitter chatter shows engagement, people paying attention to a show while watching it. Research suggests that people tweet more during a show than during advertising breaks, so marketing can relax, at least Twitter doesn’t seem to distract people from watching the great ads they’re producing for much money (although possible there is less Twitter chatter in the breaks, because people don’t pay attention to them and do something else, like rush to get a new bowl of popcorn, just saying – that would be more true for example for Germany, where there are fewer but longer breaks, but Nielsen doesn’t rate in Germany).
One thing to take into account at least at the moment is, that demographics of Twitter and of average TV audience are not the same. Twitter is younger, more urban, more tech-savvy, bit more geeky – but on the other hand, that is a group marketing is much interested in, they are the ones with money to spend (allegedly). Beyond that most of the Twitter chatter is produced by only a small number of people, many Twitter users are more listeners than talkers (although this might partially be due to craze about follower numbers proving something, so there are not just spammers around but a number of accounts just there to increase follower numbers)
At the moment they are measuring number of tweets. Not sure if that includes new-style retweets, but I guess it does, didn’t find info about that yet; very sure it doesn’t include faving a tweet. And they measure how many people theoretically are reached by those tweets, aka unique followers (taking into account, that people have partially the same followers, so it’s not a mere addition of follower numbers of the accounts which tweeted).
Don’t mistake these numbers with what you can see on the Twitter webside as trends. Twittertrends are something different, a relative measure of the impact of a subject, relative to the overall number of tweets send in a certain time frame. In other words: what is popular chat on Twitter at a given moment. Additionally the trends shown are those which are newly popular, meaning, if a subject is popular all day it will likely not be in the trends all day, only if there is a sudden significant increase in chatter again (to not bore us all to death with ongoing Bieber or Cyrus chatter every day). So a show trending (while on air) might be still far from being in the top Twitter TV rating charts, simply when there is little Twitter chatter in that time at all, and because there was very little talk about it before, but all in all it didn’t generate huge tweet numbers. And the trends depend on who you are following and on location (but you can control the latter a little in your settings) More about Twitter trends algorithm.
Now that they’re making money with the data, Nielsen and Twitter are of course not as free with giving data anymore. So haven’t seen any numbers on the web about episode Lost Girl.
¯\_(?????? ?)_/¯