Turkey is all a-Twitter: The Politics of Social Media

Last week the Turkish prime minister banned the use of Twitter – and failed most spectacularly.

Turkish residents immediately took to Twitter to air their grievances on the ban, helped by Twitter itself which issued a message to let the people of Turkey know they could SMS tweets.

Needless to say, social media and news outlets were on fire with the news, including commentary from the Turkish president who opposed the ban, and of course deep analyzation of what exactly was going on in the world of Twitter.

Brandwatch client We Are Social set us the challenge of unearthing how the Turkish public was responding to the ban.

Had every person in Turkey abruptly stopped Tweeting? Were the Turkish people bombarding Facebook and Tumblr blogs in lieu of tweeting to vent about the ban? Did anyone even care? What was going on!?

Brandwatch was on the case…

Having built the Brandwatch system for the purpose of monitoring real world events, we were keen to put our technical muscles to work and analyze relevant data to quickly determine how the ban had affected Twitter usage in Turkey.

So how did we do this?

Our Head of Data, Robert Stanley, got to the bottom of the Turkish Twitter debacle. A Brandwatch employee for four years, Rob has had every type of data quandary thrown at him! Here’s how Rob approached this set of data:

---

Geo-location, location, location!

 

In order for our data to be credible, we had to be able to verify that Twitter data was actually coming from within the country of Turkey. When you Tweet, you’re able to indicate your location in a number of ways:

• Setting a Twitter Account time zone to a specific city

• Turning on the location settings on a smartphone Twitter app

• Tagging a location when writing a Tweet

This information is captured by Twitter and allows Brandwatch to see this data through our partner Gnip. Even better, we are able to filter our searches of Twitter using this data. However, it doesn’t make sense to simply search for all the Tweets posted from a single location. Enter keywords.

---

Choose Your Words Wisely

 

We chose a random selection of commonly used Turkish words which we felt would provide us with a sample of data which was not skewed by trending topics on Twitter. You can see the selection of Turkish words we used in our query below:

(bir OR ve OR olmak OR bu OR için OR o OR ben OR demek OR çok OR yapmak OR ne OR gibi OR daha OR almak OR var OR kendi OR gelmek OR ile OR vermek OR ama OR sonra OR kadar OR yer OR en OR insan)

Roughly translated, this is the equivalent of searching Twitter for the Turkish versions of “and,” “the,” “he,” “she” and other frequently used words.

---

Timing is Everything

 

To further validate our data, we compared the volume of tweets posted in a specific time period on Wednesday, March 19 before the ban went into effect, and on Friday, March 21 immediately following the ban.

By further narrowing the parameters of our data set, we could refine our search for the most accurate results.

---

Survey Says…

 

Not a survey per se but our data results, showed an increase in volume of over 138 percent in Tweets posted from Turkey before the ban and just after. The ban was anything but, instigating a large spike in Tweet volume from Twitter users in Turkey.

Headline-grabbing social media data about the use of social media, pretty powerful stuff!

If you’re a journalist looking for social media data on a specific trend, event, breaking news (or just want to find out what people are saying online about any topic) contact me at [email protected].