What’s the most important metric for online science communicators?
TLDR; Try focussing on just one statistic that best reflects your most important goal. Comments and conversations are often the only reliable way to gauge actual science learning, and always use the information to directly inform the content you produce and share.
Sometimes it feels like the key outcome of our social media activity is to generate statistics that feed impressive-looking monthly reports. It’s easy to lose sight of the fact that those reports should reflect the actual goals of activity, not be an end-point themselves. My advice for most people trying to make sense of their social media analytics: for the next couple of months, dramatically simplify what you’re looking at.
For large organisations with huge followings, hefy budgets and complex tracking software, detailed daily analysis might make sense. But if you’re trying to make the most of a small- to medium-sized organisation’s social media presence to communicate science and build your profiles, trying to absorb all the various numbers (RTs, reach, engagement, clicks, likes, shares etc etc) at once can leave you unclear what specific action to take. You’re probably also trying to do too much with your posts, and this will help focus your goals. By taking just one metric you’ll be able to draw direct conclusions about what is and isn’t working for you. Decide your single top priority (is it new followers, or time spent with your posts, or conversations started…), and identify the metric that best reflects that.
I asked a few key people what their priorities are:
Conversions and click throughs
In our field there’s often a reluctance to admit when social media activity is almost entirely aimed at driving traffic or sales. But when this is the case, acknowledge it and really focus content strategy on click-throughs and conversions. Brice Russ who looks after social media for Science, does just this. Their goals are clear, and everything is about optimising for visits to the website. See which posts give best results in terms of click-throughs, and learn from that.
Shares
Jason Townsend at NASA sees analytics for hundreds of accounts, each with thousands or millions of followers. What’s the one key metric he’s looking for? Shares. A key goal for them is reaching more people and shares take you beyond your immediate followers and into their networks. It’s also a great measure of whether your posts are hitting the right tone with your followers. Jason compared it to mentioning something you’ve seen or read in the pub: it’s an endorsement, and shows your content is entering people’s lives.
Retention time
If you’re trying to actually communicate science and educate people with your content, particularly with video, retention time is likely your best bet. This is what Elaine Seward who works on ACS Reactions and Luke Groskin from Science Friday both prioritise. How engaged are your viewers? Are they caught in the narrative, learning at the pace you’re setting, and sticking with it? Looking at retention time, and the moments that views drop off at, is also really helpful for guiding your production: what is triggering audience loss? What’s unique about your videos that keep viewers watching?
To gauge the impact of your science communication, don’t neglect the qualitative
Ultimately, if you’re trying to communicate science on social media, there’s only so much any numbers can tell you. Don’t let your scientific pursuit of an ideal data-driven strategy scare you away from looking for evidence of learning and engagement in the comments and interactions.
Across the groups I spoke with, there was a real trend towards reducing reliance on aggregate statistics, and towards drawing meaning from more qualitative sources. They want to be shape conversations, not just shout to a larger crowd. For example, the social media team at NASA are dealing with such large numbers that they can all become meaningless. Instead they often look to measures of impact on the tone of online discourse.
As the Cassini spacecraft hurtled its final resting place on Saturn, the conversation was dominated by a sense of finality, and concluding celebration. The team kept track of this using word clouds (Personal disclosure: I’ve never been a fan of word clouds, but in this case they are just right), and worked to shift the narrative: they started focussing on all the future work still to be done with Cassini’s data, and saw the dominant content of the word clouds shift towards future-looking excitement.
Instagrammer Samantha Yammine has a unique way of probing whether her followers are engaging with her detailed posts or just looking at the pictures. She’ll intentionally hide particularly interesting facts or questions in the middle of her posts. Questions at the start or end almost always get a response, but if people are commenting on something from the heart of a post, that’s great evidence that they’re taking in the details.
Adam Cole of NPR’s Skunk Bear channel agrees, taking much more notice of comments than statistics. He wants to see evidence of people watching the videos and acting on them: ‘I watched this then showed my daughter’, for example. This sort of thing is evidence that your work has impacted someone in a way that no view count can do.
Comments are also the best indication that your posts haven’t explained what you wanted them to. If the comments reveal that your readers have got the wrong end of the stick or struggled with a concept, the chances are you haven’t explained it well enough. John Timmer, who writes for Ars Technica, says this is an invaluable source of feedback not to be sniffed at.
Rise above it
A select few people I spoke to said they have the luxury of trying to pay as little attention as possible to their stats, beyond the unavoidable view counts that sit below every video, and instead focus on creating the quality content they believe in. Obviously, we can’t all do that, but it’s worth noting that these were also some of the organisations creating the very best content. We can all benefit from using data to guide and improve what we’re doing, but whatever you do, don’t become a slave to statistics: have faith in your own judgement too.
This post forms part of the publication, ‘Communicating Science with Social Media’, which is the product of a 2017 Winston Churchill Fellowship. Read more about the project here, and for more about me, including examples of my own work, visit anthony-lewis.com.
