Mapping Emotions on Twitter: Anxiety vs. Happiness

Posted on by

A couple weeks ago, Dan Zarrella - "The Social Media Scientist" - blogged about his experiment to map emotions geographically using Twitter. He combined the linguistic analysis system called TweetPsych with Twitter's geotagging data and the Google Maps API to plot tweets based on their emotional content in and around the city of Boston.

Here's what he found:

Happiness Tweet MapSadness Tweet MapSexuality Tweet MapAnxiety Tweet Map

The data here isn't 100% accurate (Dan admits Twitter's geotagging technology isn't great yet), but none-the-less, it's still pretty interesting.

From a quick glance, there seem to be a lot of 'anxiety' dots compared to the dots on the other maps. Can anyone who's been to Boston confirm this -- are Bostoners more anxious than they are happy, sad, or sexual? (Err, maybe keep that answer to yourself...).

Joking aside, it seems to say a lot about what people are talking about via social media.

As a reputation management company, we (VendAsta) help SMBs monitor what their customers are saying about them online. Sometimes small businesses don't believe us when we tell them that people are talking about them on the internet, or that a negative tweet/review/blog post can have a crippling effect on their online reputation. But with so many people tweeting now - and, evidently, tweeting mostly about the things that make them anxious/nervous/uneasy - even SMBs need to think about what they're doing to monitor and improve their online image.

You've heard it before: in the real world, a customer who has a good experience with a company tells three people, and a customer who has a bad experience tells ten (depending on which stat you're going by). This seems to hold up in the online world, too, where more negative reviews are posted than positive ones.

But maybe, just maybe, that gap isn't as big as it used to be. There are a lot of dots on Dan's happiness map, too - almost as many as the anxiety map - and perhaps social media is evening the playing field when it comes to sharing positive and negative thoughts.

Now there's no way of telling how many dots on these maps were related to a business experience, but it's not unreasonable to assume that the distribution of causes across the emotions was pretty similar. Even within sexuality people could have been talking about companies - someone might have tweeted about how much they "love McDonalds/Wendys/KFC/etc." and got caught in the wrong category (unless, of course, they actually do have forbidden intimate feelings for Ronald/Wendy/The Colonel/etc.). Sentiment engines aren't perfect yet, but they're getting there; the day computers understand sarcasm, exaggeration, and other high-level human emotions, by the way, is the day Siri takes over the world.

I digress.

At the end of the day, this experiment is an interesting look at our current Zeitgeist. It's safe to assume that we'll be seeing more and more research like this in the future. In the mean time, start considering how your company is handling its online reputation and what you're doing to increase your customers' happiness and reduce their anxiety.


VendAsta is the market leader in online reputation management. With a full suite of white label technology solutions, we bring social media excellence to digital marketing providers including online yellow pages, newspapers, broadcasters, SEO services, Certified Marketing Representatives, and interactive agencies. For more information, visit