Subject latency on Twitter: #royalbaby

As if we needed another lesson in Twitter’s short attention span…

The birth of the royal baby gives us another good look — with a large sample –size of how people tweet and RT.

The below graph is the number of tweets with #royalbaby between 8 a.m. (CDT) and 8 p.m. July 22.

Twitter announced around 8:25 p.m. that there had been about 2 million tweets about the baby’s imminent arrival since the time Kate Middleton was admitted to the hospital.  You can see in the Topsy data for the hours before 2 p.m. CDT that the tweet interest is running in that 28,000-35,000/per hour range (there’s a slight dip between 1 p.m. and 2 p.m. where the numbers hit about 28,000/TPH and I’d like to think it’s more evidence of the short attention span, but the sample is too small. Between 2 p.m. and 3 p.m., we see a spike in volume — 293, 669 tweets.


Diving into the numbers, it looks like this:

Between 2:30 p.m. and 2:31 p.m., there are 1,795 tweets. Between 2:31 p.m. and 2:32 p.m., there were 6,268 tweets. in the next minute, there were 8,622 tweets. The minute after that (2:33 p.m. – 2:34 p.m.) there were 10,183 tweets.  Between 2:36 p.m. and 2:37 p.m., we see the peak of 11, 968 tweets (which is about the TPM the 2013 Super Bowl sustained).

But it falls off the cliff pretty rapidly from there.  At 2:53 p.m., we’re at 8,633 tweets per minute. By 3:40 p.m., we’re at 2,643 tweets per minute. Here’s a Topsy graph showing tweets per hour a couple hours after the birth announcement.


2 p.m. – 3 p.m.: 294, 122 tweets
3 p.m.  – 4 p.m.: 216,442
4 p.m. – 5 p.. : 72,259

It’s the same pattern we saw with the #zimmerman tweets: Breaking news of a singular nature — a jury verdict, a birth — that does not have ongoing developments is fairly quickly abandoned by Twitter users.