Luckily Twitter keeps a record of the tool that is used to publish each status update, so I decided to write a tool which would automatically analyse my feed.
The following table shows tools that I used for the 1,000 most recent tweets from my account (ignoring the tools which were responsible for less than 1% of the Tweets each):
Percentage | Count | Tool |
31.67% | 317 | web |
24.38% | 244 | TweetCaster for Android |
11.49% | 115 | foursquare |
10.79% | 108 | |
4.00% | 40 | Tweet Button |
3.70% | 37 | HootSuite |
3.30% | 33 | |
3.30% | 33 | Twitter for Android |
2.50% | 25 | NetworkedBlogs |
1.60% | 16 | bitly |
1.30% | 13 |
Another thing that people sometime comment upon is whether Twitter is a broadcast service or a conversational service. So I also looked at what percentage of my tweets were replies to other people (8.7%). However, almost half of my tweets (42.8%) are accounted for by retweets of content originally Tweeted by someone else.
An academic once told me that my social networking style was "to collect interesting information and share it with my friends". However, a more honest explanation would be that I am lazy and it is much easier to click on a "retweet" button than to think up original content to Tweet about.
My colleague Andy Stanford-Clark was recently featured on a list of Tech Execs that smart people follow and so I decided to compare his activity to mine. More than half of his tweets (67.5%) were replies to other people and retweets accounted for only 10.8% of his tweets. Therefore it is clear that he uses Twitter as a conversational service more than I do.
He also used a much smaller set of tools to interact with Twitter. He only used 8 different tools (as compared to my 18) and over half of all of his updates (65.4%) were done directly on the Twitter web site.
Percentage | Count | Tool |
65.40% | 654 | web |
8.80% | 88 | Twitter for iPad |
8.70% | 87 | TweetCaster for Android |
6.80% | 68 | txt |
5.40% | 54 | Twitter for Android |
4.70% | 47 | TweetChat |
0.10% | 1 | HTC Peep |
0.10% | 1 | Tweet Button |
It is also interesting that Andy is a much more prolific user of Twitter than me. I had to go back as far as 23/Nov last year (about 6 months) to get 1,000 tweets from my feed, while I only needed to 10/April this year (slightly over a month) to get 1,000 tweets from Andy's feed.
It would be interesting to a similar analysis on a wider sample of Twitter users, because I guess there is no such thing as a typical Twitter user.