Twitscoop

When you first come across Twitscoop (we found it on one of the TweetDeck channels), it passes off as yet another mildly interesting tag cloud. But digging into the application reveals a powerful real time analysis tool.

The main page provides a constantly changing tag cloud for the hottest terms on Twitter. Just watching this page and comparing how some terms have more staying power than others can easily kill an hour.

Mouseover one of the terms (e.g. Leo) and you will see the term in the context of the latest tweets.

On the right hand side of the page are “trends” – these are terms that have been hot for several days if not weeks. Hover your mouse over one of these terms and you’ll get not only the context, but also a pop-up graphic that shows the ebb and flow of the term’s popularity over time. In this case, you can see that John McCain’s appearance on Letterman generated a lot of traffic and then it quickly dissipated.

That’s the “push” part of the data from Twitscoop. Here’s how to “pull” data.

Start on the main page with the search box on the right hand side. Input a twitter username or keywords to track a conversation, topic or conference. The results will auto-refresh every 20 seconds.

We started with the term “beer” and got the results that you would probably expect on a Friday afternoon. Some people were thinking beer as early as lunch time and the momentum slowly grew until the volume had quadrupled by happy hour in the Eastern time zone.

However, what if we were trying to understand how people are thinking about a particular brand of beer on a Friday afternoon?

For the term “bud“, (we tried budweiser, but no one uses that term in twitter land…), you get an interesting distribution along with the contextual tweets – see screen shot, taken at 5:40 ET on a Friday.

If you click the link at the bottom, you’ll go to a page with more timeline options – 6 hours, 1 day or 3 days. Clicking on 3 days shows a distinct, repeatable pattern.

There are a lot of ways to use Twitscoop to monitor how terms behave in relation to time. The ability to dig into the context of the actual tweets is what makes this tool powerful.