Marketing Pilgrim's "Research" Channel

Sponsor Marketing Pilgrim's Research Channel today! Get in front of some of the most influential readers in the Internet and social media marketing industry. Contact us today!

Facebook and the Science of Word Choice


When it comes to posting on Facebook, we’ve looked at what times of day are best, which days of the week are best and whether random posts can help you market your product.

Now, Facebook has gone a step further with a scientific look at word choices on updates. In order to study the words, millions of updates were fed into a computer and broken down by their corresponding Linguistic Inquiry and Word Count (LIWC) categories. Categories referenced parts of speech (pronouns, prepositions, verbs), general topic (work, school, family) or emotional response (happy, sad, angry).

The computer then took the category numbers and sorted them by age groups, time of day and by the popularity of the account. If you want to see all the detailed charts, visit this page on Facebook. If you’ll settle for a quick overview, then stick around.

The most interesting, and somewhat sad, group of findings came about when they measured the frequency of positive and negative words throughout the course of the day. On the chart, blue is the line of happiness. As you can see, it peaks early in the morning then drops off as the day progresses when the negative updates begin to climb. Wow, what does this say for us as a society? We wake up full of hope for a great day then end up sobbing in our beer by midnight. The only saving grace here is that the pattern repeats, so we’re nothing if not resilient.

You won’t be surprised to find that younger Facebookers swear more, are more angry and they use the “I” pronoun more than older users. Older people spend more time talking about family, they use more prepositions (indicating that they write longer and more complex sentences) and they use “We” more often than “I.”

“You” belongs to the popular people, which is probably why they are. Think about it. They also talk less about the past and about their families.

Another fact that came out of this research is something we’ve always known but now have proof. Positive updates get more “likes” but negative updates get more comments. It’s just a fact of life that more people will invest their time in complaining than in complementing.

If you’re Facebooking for dollars there are a few things you can glean from the report. First off, make sure you’re using the same language as the audience you’re trying to reach. I’m not suggesting you curse in order to get teens to your fan page, but the study shows that the followers of a person tend to use the same frequency of word categories as they person they follow.

Second, is the “you” factor. Take a look at your recent status updates. How many times did you write “you” instead of “I” or “we?” Make it about your customers and they’ll be more likely to stick around.

Finally, positive versus negative. If your goal is to have more comments on your page, then controversy and negative updates will do the trick. If, however, you’re more concerned with the number of people who “like” your page, then happy and peppy is the way to start every day.

  • http://www.bsfentre.com Digital Media

    Intriguing info. Definitely going to start using “You” more! Thanks!

  • http://www.grahamjones.co.uk/ Graham Jones

    Sadly, Facebook has fooled people. The so-called correlation is nothing of the kind. The statistical data shows that the differences which are apparently demonstrated are down to chance alone. In statistical terms a positive correlation is scored as 1, which means there is a clear, positive relationship between two things. A correlation of -1 is when the two things are negatively related – when one goes up, the other goes down.

    The Facebook study has correlations from 0.2 to -0.2 – in other words, very close to zero. They have used graphics of a temperature scale to suggest visually that there is a relationship between different age groups and different word choices. But the figures prove no such thing.

    Similarly, the variation in the happy vs sad line is meaningless. The graphic makes it “look” like there is variation, but there is no scale to the axis, which means the variation could be very small indeed.

    Facebook hasn’t found what they purport have found.

    • Cynthia

      Interesting that you say this, because looking only at the graphical elements of their report, that’s how it looked to me but when I read the narrative I got a different story. I took this to mean I was misreading the graphics.

      Having said that, I don’t doubt the validity of any of the information. I work with a variety of clients on Facebook and so I see a wide range of age / popularity/ background etc. From that alone, it looks like the results are right.

      What’s really funny to me is I have a friend who updates several times a day and almost always begins his day with positive “I’m out doing great things” messages and ends with “well that didn’t turn out the way I hoped” messages. He’s this whole study in microcosm.

  • http://www.customtattoosusa.com Jen Green

    Thanks for all the intel. That’s very thorough. It’s definitely interesting how positive updates decrease throughout the day. Thanks,again.