Thursday, 8 July 2010

ReTweet relativity


One of the tools I made recently is my own Search tool that I use to make hashtag Tweet Collections. Its output is HTML but it's also valid XML (just change the extension), but the real beauty of it is that every tweet has a timestamp, including seconds - so I can tell which tweets can be human, and which must be automated

How?
That is relatively easy. Take an originating tweet, usually a blog post, allow for a few minutes in order to be able to spot the tweet, read the article and ReTweet it, et voila every tweet from then on is likely to be human. Is it the same or very similar tweet, and the ReTweet's there within seconds? Then it must be what I call a FanTweet: a tweet which, according to the Law of PEERception, simply exists because the follower is such a great fan of the Tweep him or herself that just anything is automatically ReTweeted, without reading, checking or judging it first

Today, Marshall KirkPatrick complained at exactly 20:14:52:
oh man, my bot told me about the new YouTube Mobile version *9 minutes ago* and TC & Mashable already have posts up? :(
I wondered about that so I used my Search tool and found out that not only Marshall was right, but at the time he tweeted that, the article had been ReTweeted 390 times already - albeit 150 times according to TweetMeMe

Here's the lowdown (only showing times, the date is July 7th for all):

Mashable's tweet is timestamped 20:01:22, to be immediately followed by MashableMobile and MashableVideo's tweets. The three of them use ow.ly for auto-tweeting the post

Then, there's a 60-second silence, and the storm starts. The first tweet is at 20:02:22, and within the next minute there are 140 tweets. Any of them human? One or two, apparently. The rest? Fan-tweets serviced by ff.im, url4.eu, and bit.ly, where bit.ly has 98% of all tweets

It actually isn't until 20:05:15 that the Fan-Tweets stop. Until then, there's been a pause of 0, 1 or 2 seconds between each tweet, and almost all have been identical tweets, identical URLs, and it's been just one continuous storm of ReTweets. To be exact: 357 of them.TweetMeMe at that moment shows around 140-150 tweets, so I'm unsure what they're counting. I'm counting all tweets that contain the words YouTube Revamps Mobile Experience
After that, there are actually tweets starting with "RT " to indicate they are ReTweets from other tweets

Here's my Tweet Collection of the first 2 hours of this particular Mashable post

Right now, 2 hours have passed. TweetMeMe shows 324 tweets, I count 644, Anyway, half of these are automated Fan-Tweets, and half are human tweets

My conclusion? If you're Mashable, you needn't worry about coverage, publicity and spread. But you really should worry about the ratio between ReTweets and people that actually read the content that you publish

0 reacties:

Post a Comment

Thank you for sharing your thoughts! Copy your comment before signing in...