Thursday, July 29, 2010

Is Twitter a conversational mechanism

I hear a lot of people says that Twitter is a great tool for enabling conversations. In my experience Twitter can be used for conversations, but it is far from ideal for this purpose. I think that what people are really saying is that they are surprised how much conversations happen on Twitter given that the platform was designed primarily as a broadcast mechanism (for distributing news about what you are doing).


A conversation takes place when a person reacts to what they hear someone else saying either by responding to the original speaker or by passing in the message they heard to others. In some ways, Twitter is seen as a replacement for email. In order to judge whether Twitter is more likely to be used for conversations than email I decided to compare the what percentage of the contents of my email inbox and my Twitter stream involve conversations. The results are summarised by the following table:

ChannelReTweets or ForwardsRepliesNeither
My Email Inbox
11%22%67%
My Twitter Feed 
11%4%85%

It is clear that I see much fewer conversations happening via Twitter than via email, because although 22% of the emails I receive are replies, only 4% of the Tweets that I read are replies (to my email or someone else's). It seems that both mechanisms are used equally good at spreading news because 11% of the messages on either format are forwarding an email from someone else or re-tweeting a message that someone else twittered.

These results are skewed by the fact that not all emails or Tweets come from normal people. Some of the emails I receive are from automated programs e.g. the automated software build system we use will notify when a new build is available. In a similar way, some of the Tweets appearing in my stream come from semi-automated sources e.g. the Irish Times feed is really just an automated series of links to news stories on their web site. I am not sure I am a totally normal person (if such a concept even exists), but I decided to analyse the statistical profile of what I write:

ChannelReTweets or ForwardsRepliesNeither
My Sent Emails 
14%64%22%
My Tweets39%18%43%

Although the numbers are different here, I still see that I clearly use email more for conversation than I use Twitter - the vast majority (78%) of my emails are either replies or forwards of someone else's message while only 57% of my Tweets involved either replying or forwarding someone else's message. I think that the relatively high percentage (39%) of my Tweets which are re-tweets of someone else message is an indication of the fact that I am not very confident that I know how to use this new medium and so I play it safe by re-tweeting other people's content whenever I find it interesting.

I don't claim to be an expert in advising people about the best way to use Twitter, but it is sometimes easier to identify people using the platform badly. I was amused to see the twitter feed for Poolbeg (an Irish book publisher). The feed was established on 16th of January and started off well with several tweets in the first few days, but at the end of the first week they posted a status message "Evening all - it's been a busy week for Poolbeg. Wine time now though!". Then there are no more tweets in the intervening 6 months. Clearly the person who was responsible for updating their Twitter account thought that drinking wine more interesting than Tweeting :-)

1 comment:

  1. You might like to read the article about the difference between the use of microblogging inside and outside the the IBM firewall http://bit.ly/99Me57

    ReplyDelete