The measurement and analysis of social media is still very young and far more difficult than previous examples. First, the encounters often happen through third pary systems (for example, through twitter or facebook) so measurements like bounce rate and conversion don’t make as much sense (nor are they really accessible).
More importantly, the currency is different. Web analytics and optimization are built around many individual encounters. Primarily how many single visitors react to our web page. Social media is built around connections and relationships. It’s not just how one relates to many individuals, but how one fits into the environment.
Relationships and Connections have an impact
The return on investment (ROI) is more complicated. (And, of course, for libraries the investment is less in dollars than it is in personnel.) However, there are many examples of how social media can have real returns. For example, Dell computers found that the generated $3 million in sales through Twitter. Another study found that engagement over the social web correlated to higher sales. In fact, the groups with the highest levels of engagement found an average increase of 18% over 12 months while groups with low levels of engagement found an average decline of 6%. (A pdf report of the study, which is excellent, can be found here).
Despite all of the evidence of value added, the vast majority, 84%, of social media programs do not measure the return on their programs. Not only is it hard to do, but many seem to feel it is against the spirit of the community to rank and measure engagement.
Well, fortunately we are libraries and we cannot increase profits. So consider the measurement of social media to be an exercise in strengthening our connection to patrons. After all, if people enjoy and connect to certain actions more than others shouldn’t we try to increase our efforts in those areas? So, enough intro, onto the analytics.
One of the easiest ways to analyze social media programs is to measure how our participation is accepted and used by the community. In twitter, this is often measured by two things: how many people follow our links and how often our tweets are retweeted.
There are a couple tools that can help follow these. To find statistics on links, the easiest tools are just the standard url shorteners. The most common, bit.ly, has accounts that will give statistics on how many people have clicked on the link that you’ve posted as well as how many people in total have viewed the link (for example if another users posts the shortened url).
This has two advantages, first, it helps give an idea of how “interesting” the links we post are to other users. Maybe patrons snooze over links to local election results, but love the link we post on job searching tools online. Second, we can see how links to our pages are being viewed. For example, we may see that many people have followed a link to a library program. Of course, this only measure those who follow a link through a bit.ly shortener, but they are one of the largest in the industry, so this is a minimal disadvantage.
Another way to track stats is with ow.ly. This is my favorite since it is both a twitter client (called hootsuite), a url shortener and a stats tracker in one. Here for example are my stats for the last thirty days(click for full image):
Ok, so I’ve been a little lazy the last few days. However, it is interesting to see what kinds of links my followers enjoy. Here’s a look at the top clicks:
So I can tell that many of the people following me must be librarians (no surprise there). However, they also are the type of librarians that like comics.
This is fairly low level stuff, but it’s worth considering what your patrons like.
Another tool, one that I haven’t really used, is called postrank. It combines with google analytics to measure things like blog posts and how they can take on a new life in Twitter, Reddit, FriendFeed, or other social media.
However, beyond measuring how individual posts fare. Hootsuite, along with other tools, give you the ability to track mentions of your twitter name or the ability to follow keywords. This is less important for measuring your individual participation, but can help measure discussion of the library.
This took longer than I thought. As I mentioned, this is a new area so it is very exciting and full of many different groups trying out different strategies. Next week I’ll discuss how to go beyond tracking and analyzing library participation and show tools to measure the discussion of libraries and how to follow the buzz about different library programs over twitter.