Twitter Metric Experiment Week #1

Okay, so I've been running through different kinds of comparisons using Twitter stats (see entries below) to see what can be learned about social networks based on the available data. Here is last week's data of the most popular Twitterers.

Fastest growing Twitter accounts (4/26/09-5/3/09):

1. UNHCR Refugee News (@Refugees) 5,446% SU
2. Soraya Durabi (@SorayaD) 2,424% SU
3. CDC Emergency (@CDCEmergency) 1,270% SU
4. Matthew Perry (@LangfordPerry) 1,000% 
5. NFL (@NFL) 787% SU
6. Kevin Nealon (@Kevin_Nealon) 719% SU
7. Fantastico (@ShowDavida) 457% SU
8. John Green (@RealJohnGreen) 395% SU
9. Adam Wilson (@UWBCI) 337% SU
10. MC Thumbtack (@ABigVictory) 317% SU
11. iTunes MovieTrailers (@iTunesTrailers) 302% SU
12. Mano Menezes (@ManoMenezes) 299% SU
13. The Real Paula Abdul (@PaulaAbdul) 282% SU
14. Chris Anderson (@TEDChris) 211% SU  
15. Etsy (@Etsy) 145% SU
16. Russell Simmons (@UncleRush) 135%
17. Lauren Conrad (@LCLaurenConrad) 133%
18. Lakers (@Lakers) 120% SU
19. Paul Feig (@PaulFeig) 119% SU
20. Digg 2000 (@Digg_2000) 111% SU

SU=Twitter's Suggested User

Just looking over yesterday's practice run and this list what is immediately obvious is what an impact being selected by Twitter to be put on the Suggested User list can have on an account, especially an initially small account like @Refugees which grew from 697 Followers to 38,656 in just a week. It makes a huge impact for the first & second week after an account is listed and though it will still bring accounts new Followers afterwards, the change will be less dramatic. 

It was well accepted & acknowledged that the Suggested User list (begun in February 2009) is influential in building a following on Twitter but I had no idea what a dominating influence it was. Since the data is already gathered, tomorrow I will post an "alternative" list of the fastest growing accounts that grew without the benefit of Twitter sponsorship. This list might tell a different story of what garners attention & popularity at the current moment.

As I've said, I'm not a statistician and this has been extremely time-consuming. I'm committed to running this exercise for four weeks and then will assess what, if anything, I've learned. Your contributions and thoughts are most welcome!