User21

FriendFeed has a very clean design but it doesn’t utilize wide screen real estate. On a 24″ display using 1920×1200 pixel resolution, you can show at least two columns of FF feeds and still have some room to spare.

I’m not sure how to balance width vs. scroll. Having two columns can be confusing, and I certainly don’t want to scroll down to read the left column only to scroll up again to read the right one. However I’m sure this screen real estate can be used in some way to enhance my productivity.

Maybe something along the lines of the Pagerization Greasemonkey script, only horizontal? Still thinking about it. In the meantime I’ll just keep scrolling down.

Did you know FriendFeed has two aliases, pointing at the same site?

Try these links:

• Simpli.st - this domain is registered under FriendFeed.
• Frnd.us - this domain is registered under Paul Buchheit, FriendFeed founder. It actually redirects to friendfeed.com .

I’m still fine tuning my crawler but below you’ll find the top 250 most followed users as calculated from data collected at the end of June. All previously mentioned limitations still apply. The percentage is calculated against the active public users discovered in this crawl, which is 52137 users (8143 users are private, and the current grand total is 60280).

Terminology

I changed my terminology from “active” to “most followed” as it better describes what this is all about. When I use the term “active” I mean “discoverable”, and it doesn’t matter if a person actually does anything on FF as long as his feed keeps flowing in. If she streams her twitter activity to FF but doesn’t set foot in FF, she will still be picked up by my crawler because her twitter activity contributes to her FF feed.

Margin of error or coverage limitations

Occasionally FF users mention how many followers they have. I tried comparing these numbers with my limited data and it seems to fit well within the estimated 15-20% margin.

For example, take a look at this entry and notice the numbers reported there by Susan Mernit and Jeremiah Owyang. Susan reports 1290 followers, I was able to discover 1055. Jeremiah reports 1979 followers, I was able to discover 1578.

Also, as far as the “big guys” go, I’m sure I’m missing a lot more than 15-20% which is natural given their popularity. There are probably a lot of dormant/inactive users who simply follow Scoble and Arrington without contributing anything to the feed (which makes them non-discoverable by my crawler).

The top 250

So, here is the top 250 list. As you can see, it goes “long tail” pretty fast.

I believe my little research project has gathered enough data to produce some meaningful results. Please read my previous post regarding the methodology and imitations of this analysis. Just to clarify, my analysis covers only active and public FriendFeed users, and only those I was able to discover.

But enough with the disclaimers, let’s see some numbers.

How many FriendFeed users are there anyway?

For now I stopped crawling at about 60K users since I was discovering less and less new users with every crawl until it became clear that I’m discovering users who have just joined or existing inactive users who just happened to post something and went back to sleep. Basically they don’t interest me in the context of this little research.

Factoring in a few technical limitations, API coverage issues and some secret sauce, I assume I’m missing about 15-20% of the users. In fact I could be bold and speculate that FriendFeed has an active userbase of about 75K users. Again, this is based only on partial data and my own interpretation.

What is the public/private ratio?

The total number of users my crawler discovered is 57560 .
49580 of these users are public, while 7980 are private.
So the public/private ratio is about 6:1 .

Who are the most popular FriendFeed Users?

Out of this total of 57560 users, here are the top 20 “heaviest” (most subscribed to) FF users:

You probably didn’t need my help to figure out that Robert Scoble is the most popular user on FriendFeed but my research seems to support this. Almost 20% of the public population I discovered are subscribed to his feed, 13% follow Michael Arrington, 5% follow Thomas Hawk and only 4% are “stuck following Louis Gray“…

What next?

Any suggestions for further analysis are welcome. I have my own ideas but I’m interested to know what others think and if this research has any value. Leave a comment or contact me by mail.

I wonder, maybe I should just let people search this database? You could ask “who follows louisgray” and get a list of users subscribed to him. Interesting.