Tuesday, 4 December 2012

On the insignificance of (Re)tweets to a post

In a discussion about blindly ReTweeting yesterday, I remembered that I once did a short analysis on auto-tweets.
An auto-tweet is a schedule you set up against an RSS-feed or any other trigger, which tweets the URL with a title, some of the post itself, a fixed word or hashtag, etc. Some "thought-leaders" use it to ditch all attribution and pretend the tweet originated from themselves. Usually, their Twitter bio then also says "I tweet interesting links" - well, now you know how they do it and how original and carefully curated these are...

I wondered about the current state of affairs - and did a deep analysis this time, the results of which are depicted above

As an example, I've taken a Mashable post from their Social Media / Facebook section, and counted the auto-tweets. Well, actually, I used a tool for that which searches Twitter and returns the tweets and stats for me. Use the tools, Luke! And then the proper ones of course...

On December 3 2012, 21:10:47 @mashsocialmedia tweets:

which starts off a whole load of automated, fixed, scheduled RT's. Two interesting things can be noticed: first, there are 450 tweets in the first 3 minutes, and then it dies out. My stats comprise half an hour, and we're at 14 hours later now, and the 675 tweets in the first half hour have now become 895.
If you take the first half hour and add an extra column showing incremental RT-percentage per minute (which indicates the relative division of tweets across those 30 minutes), you get the following graph which says it all:

Everything happens in the first 3 minutes: 66% of tweets that get sent in the first half hour, gets fired off then. After that, small hicks and coughs is all that remains.
Do any of those first 450 tweets involve anyone actually reading the post? Maybe 1 or 2, what the hell, maybe even 3 people. But less than 1% is a fair sugguestimation

So, all those impressive tweet stats on blogs? Half of them and more are auto-tweets by bots / schedules. Probably a quarter of the rest are from people that haven't figured out yet how these tools work, and blindly RT such a post by hand. Probably half of the remaining quarter RT's it because it's from Mashable and about Social Media and fits their agenda. That would leave 7.5% of all RT-ers that actually have thought about RT-ing this post? Sounds fair to me

There is another disturbing trend here though: the utter lack of attribution. How many tweets attribute this post to Mashable, do you think?
63. Sixty-lousy-three out of a total of 675, that is 9.3% - that means that 90.7% pretends this to be their own smart find, in stead of something they highly likely read on Twitter.
What is the attribution percentage of the first 450 guaranteed-to-be-autotweets? Well?
31. Thirty-one out of a grand total of four hundred and fifty. That is 6.9%, rounded, meaning that all these auto-tweets really put in an effort to make this appear as their own smart find.
Then, what about the currently remaining 220 that actually consist of real live people that read the post (I hope)?
52! FIFTY-TWO! Yay. That is a whopping 23.6% of real live breathing people that attributes this post to Mashable

Well well well. Is Twitter mainstream? Undeniably so, at some points it's getting or already gotten as ugly as real life. Should you stand in awe for hundreds, even thousands of tweets on a post, and (over)value the author or blog for that? Just like you should stand in awe for the big brand new car of your neighbour...

0 reacties:

Post a Comment

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