What Twitter Told Us About ILI 2011

Thoughts on #ILI2011

As I said to one of the two video bloggers who recorded participants’ thoughts and comments about the Internet Librarian International (ILI) conference, ILI is probably my favourite conference as it provides an opportunity to catch up in developments in the online Library world in the UK, Europe, North America and Australasia.  This year at ILI 2011 I could only attend for the first day, but this did give me an opportunity to hear about, amongst other topics, JISC-funded developments in the areas of usage data, analysis techniques which can help to prove value and three cutting-edge technology developments taking place in Norway, Belgium and the USA.

Unfortunately I don’t have the time to give detailed thoughts on the sessions I attended. However an analysis of Twitter usage at the conference might help to provide some insights into how Twitter was used at the conference.

What Does Twitter Tell Us?

If you carry out a sentiment analysis of the archive of the tweets from last week’s #ili2011 (Internet Librarian International) conference I suspect you’ll find a lot of positive comments.  Without going into a textual analysis of the content, what can we learn from the Summarzr statistics of the 2,683 tweets from 310 users? (Note as described in a post on Conventions For Metrics For Event-Related Tweets I feel that such summaries should include a data range, so this total covers the period from slightly after the start of the opening plenary talk on Thursday 27 October at 08:38 (actually 09:38) to Saturday 29 October at 09:37).

As perhaps might be expected for an event with over 300 librarians and information professions the Twitter users understood the benefits of providing distinct tags for the three parallel streams. This is a bit of a hobbyhorse for me and I was pleased that I was able to set a precedent in the first set of parallel sessions when I encouraged the 100 or so participants in the session on “A101 – What’s on the Technology Horizon?” to use the tag #A101 to be able to differentiate the conversation from those taking part in sessions “B101 – Not So Secret Weapons – Advocacy and Influence” and “C101 – The e-Book Revolution in Libraries“:

ili2011 (2676) , a101 (98), c202 (67), lidp (61), a104 (54), a203 (53), b103 (45), a102 (44), a201 (41) and b202 (32).

The easily-identifiable tweets will help myself and Åke Nygren, my fellow speaker in the session, to be able to see what was being discussed during our talk, so such session tagging provides a useful way for speakers to gain feedback for their talks. Our opening track seems to have been the only one in which significant numbers of session-tagged tweets were used. However it seems that the benefits of such tagging were quickly spotted with the second, third and fourth parallel sessions (which end in 2, 3 and 4) being included in the above list of the top ten hashtag contained in the TwapperKeeper archive. I should also add that in revisiting my post on Thoughts on ILI 2010 it seems that use of session hashtags is new this year, with only session #C102 being included in the list of top ten hashtag for last year’s event. (Having just looked at last year’s programme it seems that Session C102 on Monitoring and Maximising Organisational Impact was given by myself and Joy Palmar, so it seems it has taken a year for this practice to become embedded!)

The list of the top Twitterers at the conference included several of the ‘normal suspects’ who have a proven track record of tweeting at conferences headed, as was the case for ILI 2010 by @bethanar and @Mimomurr.

Comparing the overall numbers of tweets at the year’s events with ILI 2010 it seems that Twitter usage has now stabilised:

ILI 2011: 80% (2150) of the tweets in this TwapperKeeper archive were made by 14% (45) of the twitterers. The top 10 (3%) twitterers account for 46% (1241) of the tweets. 56% (175) of the twitterers only tweeted once.

ILI 2010: 80% (2032) of the tweets in this TwapperKeeper archive were made by 15% (57) of the twitterers. The top 10 (2%) twitterers account for 45% (1143) of the tweets. 61% (229) of the twitterers only tweeted once.

It should also be noted that once again there were very few geo-located tweets: 39 tweets this year compared with 18 last year, both of which represent no more than 1% of the total number of tweets.

Feedback From Twitter

The event organisers have sent out a SurveyMonkey form to ILI 2011 participants which will help to inform planning for next year’s events. But in addition the event organisers will also be able to analyse the content of the tweets.

I have created a Storify page which summarises a number of tweets related to particants’ thoughts on the conference, rather than comments on the topics discussed at the conference. The most recent tweets are shown in the accompanying screen shot.

Beyond ILI 2011

We were told about changes in ILI conference organisation, with next year’s event being the responsibility of Information Today’s office based in Oxford.  Although I’ve enjoyed previous ILIs, I do feel it will be beneficial to have greater participation from the UK and mainland Europe. I felt that it was somewhat strange, for example, that although there was much interest in use of social media, there was little discussion about privacy issues and the implications of EU privacy legislation related to cookie use.

In light of the changes to the event organisation I would like to conclude by making some suggestions related to use of social media at the event, based on the ideas I’ve described in this post which I hope with be useful for other event organisers.

• Create a TwapperKeeper (or equivalent) archive of event tweets well in advance of the conference. Note that I discovered that a TwapperKeeper archive hadn’t been set up for the #ILI2011 tweets during the opening talk. I created an archive  during the talk, but this meant that tweets made in the run-up to the event will not be included in the archive
• Be aware of the benefits of session-related (or room-related) hashtags for parallel sessions and ensure that you clearly publicise such hashtags if you wish to encourage their use.
• Be aware of  how tweets can be used in the evaluation of an event.

Finally I’d also suggest that event organisers should consider being pro-active in promoting use of the Lanyrd service. It was suggested that participants badges should include their Twitter ID. But in addition, the  Lanyrd page for ILI2011 provides an electronic means for participants to develop their professional network.  No fewer than 24 of the speakers at the conference are listed on Lanyrd, but as there are only an overall total of 42 participants, this means that the majority of the 311 people who tweeted (or the 136 who tweeted more than once) aren’t included in this network.  I think that’s a shame, as I’m a great fan of Lanyrd and have included details of my talk on the Lanyrd page. But that should be the topic of another post!