The biggest PR/marketing story of the week so far involves Coca-Cola‘s surprising announcement that social media “buzz” doesn’t translate to short-term boosts in sales. On the television front, however, research comes to the opposite conclusion: yesterday we learned, via our sister blog Lost Remote, that Nielsen finally released a yearlong study firmly tying Twitter mentions to increased ratings for popular shows.

How does that relationship work? Let’s check out the numbers:

The Twitter effect is least influential on season premieres. an 8.5% increase in buzz (or related tweets by volume) leads to a 1% bounce in viewership among the 18-34 set, while a 14% increase creates the same gains among viewers aged 35-49. For midseason episodes, however, the numbers are more impressive: the amount of buzz required to create a similar 1% ratings bump is almost half the size for episodes airing midway through the season.

Nielsen doesn’t have the numbers on season finales yet, but researchers expect the correlation there to be even stronger. We get it: a deluge of enthusiastic tweets could encourage viewers who may not be quite as interested in a show to tune back in for the season’s big cliffhanger or resolution.

It’s true that Nielsen has an interest in promoting Twitter: last year the two joined forces to create the Nielsen TV Twitter ratings, which track the most buzzed-about shows. But the reason they did this is that Nielsen realized the promotional potential of everyone’s favorite super-short attention span service.

The big point here: use Twitter as much as possible to push your TV clients or shows you love. But you already knew that, right? Good. Now everyone follow Retta, better known as Donna from NBC’s Parks and Recreation, for her great TV recaps and live tweets.

Oh, and Community will probably still get the axe. We love you, Joel McHale, but this season just isn’t very funny.