Inside View from Ireland: Tweetrush Tells Me

AN IRISH-POWERED TWITTER stats package called Tweetrush [2] gives first-generation snapshot data of usage across Twitter so I started probing it to see what I might discover about the electronic water cooler preferred by some early adopters in Ireland. In the graphic above, it's easy to see when I don't tweet. The lower bars on the graph are UTC times I'm reading, speaking or working face-to-face with people. In one quick glance, I can tell when my most productive hours occur and they happen to be the times when I'm less active on Twitter, a site that gets no fewer than 30 tweets from me on an average day. Tweetrush data has encouraged me to throttle back on those tweets, just in time for the start of a new academic year. After nearly 8000 tweets, I've concluded that I can get greater value per page load in my newsfeed aggregator instead of listening to Twitter. However, I get better tips on breaking news from Twitter, but only if it lands on my mobile phone. The mobile dimension is very important for me, since I'm untethered from my desk for most of the day as I try to fit into Twitter.

Some interesting interpretations can be read in the Tweetrush data.

Twitter is not as popular as its press coverage. User numbers in the millions often get batted around by the press when covering Twitter, but the data tracked by Tweetrush shows an active online population well below multi-million. Some of those vagrant twits can be ruled as spammers that Twitter's algorithms have trapped and quarantined. A lot of tweets could be coming from people who have made themselves "private" and it's fair to assume that plenty of tweeting is direct messaging, something not captured by Tweetrush. However, there's also a possibility that there's a natural Twitter Fade that hits subscribers as they move onto different things. Online allegiance is difficult to hold and lower Twitter numbers bear out that fact. [1]

Tweetrush Suggests Best Times to Attract Attention. A double derivative of Tweetrush data documented the best times to use Twitter to get attention of a specific geographic area. I think those times would be when the flow of tweets is lowest but when the on-screen audience is still attentive. This metric is not as simple as it sounds, because you need to get some bounce from things you tweet and that bounce often comes in other people retweeting your well-formed 140 character item. Some of those tweets and retweets result in mainstream press coverage. [4]

Tweetrush Data May Shortchange Conversation. My analysis of this potential under-counting is limited, but it looks like Tweetrush counts tweets, not replies. So anything with the @ sign in it won't be tallied as a tweet. I suspect Tweetrush doesn't count private tweets either. Those are my first impressions, and I haven't verified this finding over a day's worth of tweeting yet.

Tweetrush is Sweet Infoporn. I like stats. I like comparing like to like. And that means I like the infoporn hit that I get when clicking through Tweetstats. Tweetrush is powered by a real analytics engine using PHP and the Zend engine. It's displaying only the tip of an iceberg of live data, and it's a form of sousveillence [3] that unsuspecting twits might not fully comprehend. Recommended.

0. Walter Wynne -- "Tweetrush Launches"
1. Newsgang -- Tweetrush Shows Lower Twitter Usage Than Expected.
2. Crunchbase -- Tweetrush Profile.
3. Dennis Howlett -- "Tweetrush for Enterprise?"
4. John Collins, assistant business editor of The Irish Times -- "There was some surprise the other day when i told them i had got three exclusive stories by watching Twitter."
Previously: "Trying to fit into Twitter"