Positivity works, even on Twitter
I love data-driven insights and I came across an analysis of words/phrases that correlate between those Twitter users with a high follower count or those with a low one.
Themos Kalafatis used text mining and data mining techniques with some noteable patterns on how certain words affect your Twitter “popularity”, some examples are below:
- Hate and bed correlate with accounts with low popularity
- Sleep, Hate, Damn, Feeling, Homework, Boring, Stuck were associated with accounts with very low follower counts
- Thank you is highly related to a user’s high popularity
- Student and Engineer in user bios appear to correlate with low popularity accounts
- Whereas popular accounts have words like CEO, social, media, marketing, founder, author, entrepeneur, internet in their user bios
Themos’ blog, including his methodology and data for the examples above, can be found here.
2 Responses to “Positivity works, even on Twitter”
May 20th, 2009 saat: 10:45 am
But… but… I’m an engineering student! What to do, what to do?
Interesting correlations though
May 20th, 2009 saat: 6:41 pm
hehe, just say “Thank you” a lot
In all seriousness, his older entries have some “must haves” if you want to build your Twitter audience - there are also dozens and dozens of how-tos online.
Leave a Reply