This week I’m going to base my post on some interesting research by Torgeir Aleti, Paul Harrington, Marc Cheong and Will Turner that looks at how the Australian brewing industry influence consumers on Twitter.
The research is fairly comprehensive and is structured around proving, or disproving a number of hypothesis based on a series of constructs;
• Indegree influence and retweet influence
• Message formulation and language
• Native platform behaviour
• Reciprocity and;
One of the key points that the authors make is that organisations on Social Media should be using maven-like behaviour to influence their audiences (Feick and Price 1987). In Malcolm Gladwell’s 2000 book The Tipping Point Gladwell defines Mavens as “Those who influence through their ability to accumulate, interpret, and share information” and I certainly agree that mavens can stimulate and encourage engagement on Twitter especially with audiences seeking (and wanting to share) information.
What I found really interesting in the approach the authors took was the acknowledgement that the traditional influencers (opinion leaders and highly-connected individuals etc.) may not actually be as influential when it came down to how small businesses interact on social media and that ordinary users could be just as influential, if not more so. This makes it even more important to create the right information that people are seeking and would be likely to want to share.
So the key to this it would appear, as is the case for most things in Social Media, is content. And as Bill Gates rightly proclaimed in 1996 “Content is King”. I’m sure there is a quote out there somewhere that will say “all roads lead to content”, if there isn’t, I’ll claim it now because pretty much everything on Social Media in some shape or form seems to come back to that.
Without going through Aleti’s paper in its entirety, there are a couple of areas relating to Social Media content that pricked my attention. The first was the hypothesis based around the frequent use of pictures and videos in tweets and the research on the use of question marks and exclamation marks being associated with higher levels of indegree and retweet influence.
The results showed that the hypothesis which was tested around more frequent use of pictures and videos being associated with higher levels of indegree and retweet influence was only partly supported by the research.
This is supported by my own research in to the use of images and video in tweets, where tweets containing images often achieve significantly higher levels of engagement than just text based tweets. Although saying that, there are always exceptions to the rule. With weather services we often have warnings that are heavily retweeted from text only tweets, but as a broad benchmark, in terms of engagement and specifically retweets; from the top 100 tweets we send out in any given week 90% will contain an image / animation or video.
And finally, the hypothesis relating to the use of question marks and exclamation marks being associated with higher levels of indegree and retweet influence was not supported by the research, which does tend to make sense. Although it was interesting that exclamation marks could be viewed as shouting and were unlikely to be retweeted at the same frequency as tweets not containing them (I think that needs more research as I suspect it could change based on certain demographics, such as age / culture). Again, there are always exceptions to the rule and like most things relating to content, it comes back to context. Here is a tweet from Louis Tomlinson from One Direction to their fans containing exclamation marks which attracted over 400,000 retweets.
And on that note, that’s a pretty big number, but what does it actually mean? Well you will just have to wait until next week’s post on “Analytics and metrics, it’s all in the interpretation” which addresses that very question.