The first meaty bit of data I’ve pulled out of the results is a timeline of the growth activity including retweets, comments and clicks through to the page. I’ve also marked the times when popular users retweeted and when it first hit the “Trending Topics” list.
And here are some rough numbers based on the first day of the experiment (after which activity slowed to a trickle):
| Total | % of Clicks | % of Comments | % of Tweets | |
|---|---|---|---|---|
| % columns show transmission rates (ie % of people exposed to page who Tweeted it) | ||||
| Clicks | 1,826 | N/A | 13.75% | 15% |
| Comments | 251 | 13.75% | N/A | 91.6% |
| Tweets | 274 | 15% | 91.6% | N/A |
I’ll be pulling more complex data out of the results also, including an attempt at actually mapping the spread of the Viral Test Tweet through Tweets, followers and retweets. And as before, if you can think of any other data points I may be missing, please let me know.
If you liked this post, don't forget to subscribe to my RSS feed or my email newsletter so you never miss the science.
October 8th, 2008 at 9:50 am
I link to my video in respoinse to your tweet would be great in.
RELATED / YOU MIGHT FIND THESE INTERESTING
Here is the link…
http://www.youtube.com/watch?v=f1-jT8A5GHQ
Also I did a blog post regarding your test here….
http://www.paradoxmarketingsolutions.com/blog/blogger.html
Look forward to hearing from you.
October 8th, 2008 at 9:55 am
I’m not sure a similar test would not lead to very different results. Statistically, you need more data — and similar examples agree such spreading is hard to model, or predict. I would not try to over-interpret it all, but you have interesting correlation elements. People do what you ask; your experiment had the usual 1/6 ratio of response.
October 8th, 2008 at 9:59 am
@bertil I agree the sample set was very small and clearly transmission rates could vary widely based on the actual message being sent.