Welcome to WebmasterWorld Guest from 126.96.36.199
Forum Moderators: rogerd
A tiny fraction of those who use the fast-growing social network phenomenon Twitter generate nearly all the content, a Harvard study shows.
That makes it hard for companies to use the micro-blogging site as an accurate gauge of public opinion, the Harvard Business School study showed. The Harvard study examined public entries of a randomly selected group of 300,000 Twitter users. The researchers studied in May the content created in the lifetime of the users' Twitter accounts.
It found that 10 percent of Twitter users generated more than 90 percent of the content, said Mikolaj Jan Piskorski, who led the research. More than half of all Twitter users post messages on the site less than once every 74 days.
Here goes a billion $ valuation down the toilet.
[edited by: Kufu at 4:40 pm (utc) on June 8, 2009]
Regardless, can't we say the same for a lot of online communities though? Where a small group of devotees create most of the content.
I guess the difference is that these communities don't get hypped out to be the next big thing, whereas as Twitter did.
It's like these guys that tried to sell flowers online before the .com bust.
...can't we say the same for a lot of online communities though? Where a small group of devotees create most of the content.
There is a massive difference. The majority of 'content' generated on Twitter is spam, where as the other communities you referred to have legitimate content created by that small group.
So far it is just a communication model and I keep deleting people who shout too much on twitter. Twitter is supposed to be used by closed friends not marketers.
I am more interested in my friends updates than the marketing messages (that I can get anywhere and everywhere)
This paper is a bit heavy going, but this paper from Carnegie Mellon delves into different roles:
sign me @notwit which you can split two different ways. :)
I've been doing quite a bit of research on the Twitter platform and it's users. I track about 10 high volume users via various methods to see what is happening. The automation on Twitter is where the inflation of statistics comes from. Once you reach a certain level of Followers on Twitter, it goes into auto-pilot. You can set up a script to auto-follow everyone who follows you and even refine it to avoid some of the crap that plagues Twitter. Once those auto-follow routines take place within high volume user accounts, what do you think happens? It becomes one big auto-follow cluster you know what!
I've been reading all the recent Harvard studies in addition to others that have spawned. Twitter Press has turned around recently and is now leaning towards the negative side. These past two weeks have been extremely challenging for them as they are battling spam to the nth degree. I see Twitter FAIL throughout prime time hours which tells me they don't have the resources available to handle the load at peak times. Fail Whale is still seen at intervals throughout the day. In 2009 May, according to Pingdom, Twitter had an Average Response Time of 1025.41 ms which of course translates into latency across a variety of functions. It can be a bear sometimes. I just sign off at those times and find something more constructive to do.
How do I feel about Twitter today? I still enjoy it immensely since I probably have many of those 10%'s in my Timeline. You get out of it what you put into it. Really. And, only 1 out of 10 people are going to get it right now. I like those kind of numbers. Now, if Twitter can get a handle on the spam and implement a few more restrictions, they may be able to sustain. If I were them, I'd surely take the next large $$$ offer and run. Before Google Wave launches. Same goes for FB and all the others. Take whatever large $$$ offers come your way before the Wave hits. After the Wave, you WILL become deprecated. Federation Protocol says so. :)
In a normal distribution (it's a bell-curve) 80% of output tends to come from about 20% of inputs - actually ish mentioned this above, referencing the Pareto Principle.
(And, for that matter, 64% of output tends to come from about 4% of input).
So, if right now, 90% of tweets come from 10% of tweeters, that's not far above what anyone would normally expect. And the figure will probably tend towards the norm over the next 24 months.
It amazes me that stats like this still surprise anyone - they don't just apply to Twitter - they apply to everything (TV viewing figures, university graduate earners, sports(wo)men and their achievements in a given sport etc.).
Meanwhile there are things like tweet deck that let you follow the the few people you really want to follow and ignore the rest. By the rest I mean all the people who are following so you will follow them and you are following them so they will follow you. All the get rich quick on the web scammers have their wanna get rich buyers getting on twitter to make money.
So again I ask. Who is reading those tweets? Is anyone out there?
This is not to criticize the people who really do use twitter to connect with a few friends or follow some top tweeters who are really offering information.