JSTOR Plant Science & Social Media: Measuring Engagement and Impact
Disclaimer: I am not a scientist (information or otherwise)
Disclaimer #1: Although this post isn’t about plants, plant science, or anything remotely botanical, I thought it might be interesting enough for those who are trying to start, maintain, or justify their social media presence for their respective services. At JSTOR Plant Science, we run about five social media outlets (Twitter, Vimeo, blog, Flickr, and DISQUS-I am classifying DISQUS as social even though it is essentially a commenting system) and this requires time more than anything. We draft, edit, schedule, track, and evaluate everything we post and, as an organization, we are looking for some ‘return’ on that social media investment. In this instance, ‘return’ doesn’t mean anything commercial, but rather a greater level of engagement with the site, the materials, and the community that uses these materials. That is a long disclaimer, but essentially I wanted to let everyone know this is a post about social media.
Disclaimer #2: I will be revealing in this post the simplistic approach I take to measuring the value of the social media services I run for JSTOR Plant Science, the site, the 1.5 million type specimens, reference works, art, and other primary sources, and the community. I have written about our use of DISQUS for sourcing data correction and resource improvement (mostly a healthy experimentation with trial and error), but over the last few years a certain pattern has emerged that I consider promising. While this data and these patterns might be promising, my use of charts is not so please bear with me. End of disclaimer.
What is ‘meaningful’ social activity around academic content?
The questions many of us involved in social media have, especially at the organizational level, are generally geared towards resource allocation and impact. Is it worth it and what is ‘it’ exactly? Time is a precious commodity and so this needs to generate an observable phenomena. Logic like this (and organizationally, this is extremely welcome logic) drives decisions. But in this instance, we have a hybrid breed of organizational logic applied to academic investigation. How do we measure impact for services geared towards academic and scientific practice?
In the case of the plant sciences database, we (for lack of imagination) identified the scientific process being practiced (collect plants, identify plants, classify plants, preserve plants), and latched onto that practice. We use it as a guide for what is relevant to our community, what adds value, and what might be useful for their practice. So, our social policy is driven towards
- data correction and updating (on the site)
- identifying plants and debating those identifications (on the site)
- education-telling the word what these scientists are doing and why it is important
So the above guides our posts and communication, it is often difficult to determine what demonstrates that social media is having an impact (even if that impact is greater exposure to the work of the community). I think different metrics beyond the simple visits, page views, and followers is needed to demonstrate meaningful social activity. Although I have included visits and page views, I am more curious about the social activity that surrounds these content accesses. Some of that data is elusive (for me) and many of the measures I will be including in the subsequent graphs are apple/orange comparisons. But it is a start.
In the chart below, we see site use and social media interaction over three different time periods (these are monthly averages more or less). While I shouldn’t have presented site vists and page views together (only one is necessary, in my estimation), I am interested in the general uptick of social media use over those same times. While all metrics are increasing, the social media metrics are growing as a percentage of overall use. That seems promising to me.
Working Assumption: Social Activity as Percentage of Overall Activity is measure of user investment in the resource
So basically we have two different kinds of data, the site data (usage of the Plant Science site and materials) and then the social activity around that site and those materials. Two interwoven, symbiotic bits of data, but distinct enough to drawn attention to.
So, my working assumption was that social activity as percentage of total activity would be a good working metric to determine impact. So for the purposes of JSTOR Plant Science, we are hovering around 8% of social traffic as percentage of overall activity. We have modest visits and page views for the site overall (around 35000 visits and 100,000+ pageviews per month). The social media use is around 8000-10000 ‘interactions’ per month. So, we are looking at about 110000 ‘interactions’ on a good month.
As for the DISQUS and Twitter traffic listed below, this represents ‘significant interaction’ in my estimation as it is a measure of trust in sharing content (Twitter, retweeting or otherwise); for DISQUS the comments generally represent an investment in improving the site or the plant materials contained therein (towards meeting the scientific goals of the community overall). As most of the comments are directed towards data correction or plant identification (and some genealogical research), I believe this stands true.
What doesn’t really work is the inclusion of DISQUS and Twitter on this chart as it equates all of this activity as equal. It isn’t, but I just didn’t know where else to put it. Over the last year we have seen steady increases in traffic overall (for both the site and the social media services) and an increase in the percentage of social activity vs. overall use. I believe this represents a positive investment from our audience in using the site and, I believe, warrants the resources our organization dedicated to it. We are most curious to hear how others are measuring the value and impact of their efforts in social media so please let us know and we would be happy to start a dialogue.