I might have been a little provocative in my blog-post about 'nonsense numbers' in the tracking of online reach and engagement, but the reactions suggest that some people at least agreed with my cynicism. In this post I summarise a couple of posts from from my colleague, Pier Andrea Pirani, that address the crucial issue of tracking engagement. The blogs speak quite a lot of geek, hence the summary. The details will be useful to communications specialists but also, I believe, to a more general audience as they illustrate the power of systematic attention to signposting content through the use of tags - classification terms - and of building your social media network of connections.
There were some useful reactions to the original post, on Twitter of course, including from Andreas Sandre, who added Topsy to the list of useful monitoring sites. "All true, but why throw baby out with the bath water? Gd to hv broad mixture of data including Imprssns/Accnts Rchd.... but of course, making clear the limitations of data always essential", tweeted Marek Zaremba Pike, adding that interpretation, of course, depends on whether you're trying to reach a broad audience. In that case "LikelyViews" still might be relevant. Pirani started from the same place, noting that big numbers are not just 'vanity metrics' but, taken in the round, serve a basic reporting purpose, a quick way to highlight activity levels for busy managers. And the numbers showing activity around the ICT4Ag conference kept rising, both in terms of the putative reach - now more than 2 million - but also of an impressive level of overall activity - 11,900 posts contributed by 1,272 users.
In the first of two posts, Pirani points out that his chosen tool, Keyhole, provides some useful indicators that go beyond the pure headline numbers. "It gives you insights on the contents that are shared (at the level of domains and individual links). It also provides a useful map that shows where the conversations around the hashtag are happening, the demographics of contributors and the share of posts between original posts, retweets and replies." Of more interest was his work on using tags - classification terms - to track content and interaction. He had discovered a tool by Martin Hawksey which both archives tweets - essential for later analysis, since tweets disapear from Twitter relatively quickly - and visualises interaction.
We agreed a tagging structure for the conference, ensuring that social reporters and others commenting from the event followed a system that mapped their output to the different sessions. Pirani explains fully in his second blog post how Hawksey's tool pulls the Twitter archive, drawn from the conference tag - #ICT4Ag13 - into a Google spreadsheet.This is the data for the interactive visualization of the conversations on Twitter, illustrated above. The tool maps replies, retweets and mentions, and the relations between different users. It also provides a complete Twitter archive that can be easily browsed and searched by keyword or by username.
As well as providing useful data, the results illustrate, if anyone still has any doubt, that connecting and talking to people on social media generates traffic and attracts attention, especially if you're careful about quickly classifying - tagging - what you're doing.