Last week we had our own most “viral” moment to date, and it was such an intense and fascinating experience that we thought we should document it.
September 18th saw the Scottish Referendum finally come to vote. We had been watching the build-up on Twitter for some time so knew that the #Yes campaign was significantly more vocal, at least in Scotland, which was very much at odds with the Polls.
That same week we had added some new functionality to our visualisation tool which enabled displaying words rather than just bubbles, so we thought it could be a good opportunity to see how that looked for the #IndyRef.
We created a visualisation, focused on the UK, which illustrated the strengths of where Yes and No were most popular. We used the three most popular hashtags for each camp – #voteyes, #yesscotland, #voteyesscotland for Yes, and #voteno, #bettertogether, #nothanks for No.
You can see the visualisation here.
Sure enough it was very visually striking, particularly because how geographically diverse the Yes and No camps were. Scotland looked to be all Yes and England all No. It was also interesting to see so many Yes votes from around the world.
Words also worked well for this visualisation because they were short, and in this context it felt more emotive than circles.
We had also been working on the creation of animated GIFs for our visualisations so thought it a good opportunity to see how much they increased the visibility of our tweets.
The earliest retweets and mentions were from passionate supporters from the Yes campaign, because it did appear to show a lot of support for them.
Many journalists and news organisations also helped spread it, because it did look like an interesting angle on a huge story. In the end we saw that it was embedded or linked to by sites such as The Scotsman, The Guardian, Daily Telegraph, El Pais, The Age, Zeit, El Pais, Buzzfeed, Mashable, Vox, Huffington Post, and Daily Kos.
Through Google Analytics and our site logs we could see how thousands of people were watching it simultaneously. Through Twitter’s own analytics we could see the tweet engagement, and we were tracking mentions of Trendsmap and the url of the visualisation through Twitter’s advanced search. We also used our own Trendsmap analytics to get more detail on who the biggest sharers were, where they were, and what words were being used to describe it. Fortunately it was mostly “cool”.
It became quite meta when the tweets about the visualisation started appearing in the top tweets within the visualisation. This was even more of an achievement because we didn’t even use the same hashtags in our original tweet (#IndyRef) as those being tracked in the visualisation.
We then started getting tweets that Trendsmap itself was trending, from our trending alert accounts.
When we got up a few hours later we saw that it had received over almost 200,000 views and was still going strong. Gradually interest faded as the results started to come in, which of course told a different story. When the final results were announced we then saw that the No mentions did in the come out on top in the end, and it becomes quite a poignant end to such a passionate campaign, as the No mentions finally mute the Yes.
And of course, there is a lesson that you cannot boil down the votes across a country to a few hashtags. Putting aside how representative Twitter can be, any group seeking change (Yes) will need to shout louder than those who were content to keep things going as they were (No).
Thanks for getting to the end of the story. We hope to have another similar story soon! If you have any ideas of visualisations you would like to see, or if you would like to test the visualisation tool when it is available, please get in touch.