A colleague drew my attention to Alice Marwick’s work on social media influence, and, while I was searching on the topic, I came across the video of a really interesting talk that she delivered at the Berkman Center for Internet and Society. In this talk, Alice reports on a 3-year study of Silicon Valley entrepreneurs and Internet celebrities.
In the video, Alice states that the currency on web 2.0 is ‘social status’, which she defines as the amount of attention and visibility that someone commands online. She further notes that all leading social media platforms have some way of signalling a user’s social status. So, for instance, Twitter displays the number of followers very visibly, Foursquare has a leader board and Yelp awards an elite badge.
Around minute 11, Alice makes two really interesting observations. The first one is that the reward system favours some patterns of socialisation over others. For instance, the foursquare example mentioned around minute 11.32 rewards users that go out regularly, with many different people, to different places, in an urban environment. So, users that are loyal to a particular establishment or socialise at home are less likely to acquire social status in this application.
[UPDATE / CORRECTION: Reader Arjan Tupan corrected me on how you acquire points on Foursquare. As you can see in the comments to this post, he explains that: “Foursquare also awards bonus points for checking in at the same location 3 times in a week, or for checking in with the same friend 3 times in a week, and every time in that week after that. And then there is the mayor bonus, plus a bonus for hanging out with the mayor. So, sitting at home with your partner can get you both many points on Foursquare, as long as your home is a venue, you both check in and one of you is the mayor.”]
The second interesting observation is that the way in which social status is measured and signalled in each of these communities has a prescriptive effect on users. That is, it shapes how users behave in that community and interact with each other. For instance, awarding rewards for answering questions encourages users to do exactly that.
In summary, the design of a social status signalling system has two key consequences:
- It motivates some users to participate, but pushes others out;
- It encourages some behaviours while silencing others.
These findings mean that academics studying online influence or designing experimental studies need to consider very carefully whether the users participating in a particular community are the ones relevant for the study. They also need to take into account how the observations may be biased; in particular, whether the behaviours present / absent result from a perverse consequence of the signalling system.
In turn, marketing managers trying to identify or engage with social media influencers need to assess whether the social status signals are attracting the desired type of influencer, i.e. the one that resonates with their target customers. And they need to ensure that the influencer’s motivation is aligned with the marketing goals – for instance, a foursquare user scoring high on the leader board may have a high kudos among the foursquare community, but may not be a particularly loyal customer, or indeed a high spending one. There is so much more to online influence than a number or a badge!
You can learn more about this research project in Alice Marwick’s book, ‘Status Update – Celebrity, Publicity, and Branding in the Social Media Age’. I can’t wait to read it! In the meantime, watch the talk below and let me know what you think:
Ana Canhoto 
Interesting, Ana, as always. However, I have some issue with the statements that you ascribe to Alice Marwick. I did not completely watch the video, so I’m going on your word here, but I think the basic assumption that indicators like followers on Twitter or the leaderboard on Foursquare equate to influence is false. I think number of tweets retweeted, favorited or appearances on lists are better indicators of potential influence on Twitter than number of followers.
As for foursquare, I watched that fragment around 11m30 in the video, and it’s a conclusion she draws, based on too small a sample. First of all, in the example she shows, the user in question gets points for returning to a certain location, and checking in with a friend. Foursquare also awards bonus points for checking in at the same location 3 times in a week, or for checking in with the same friend 3 times in a week, and every time in that week after that. And then there is the mayor bonus, plus a bonus for hanging out with the mayor. So, sitting at home with your partner can get you both many points on Foursquare, as long as your home is a venue, you both check in and one of you is the mayor.
But, more importantly, the leaderboard is not the main indicator of influence on foursquare. The main indicator would be how many people like the tips you left at venues, or how many people follow the lists you have created with venues to visit. I assume, but am not sure, that this is also the basis on which foursquare shows tips on venues when you check in. So, the more liked your tip is, the more likely it is to be shown to users checking in on that venue, and the more likely it is that they follow that tips of yours. (An example of a tip for a restaurant could be: “take the sauerkraut, it’s delicious.”) Now, that, in my opinion, is more likely to be real influence, than being on top of the leaderboard.
Hmm, sorry, a bit of a long response. But it’s a very interesting topic. I guess that what I as an enthusiastic user of several social networks would love to see academics come up with, is a system that measures real world influence on social networks, and that goes beyond the number of followers…
LikeLike
Thank you very much Arjan for both the time that you took and for the additional information about foursquare. No need to apologise – on the contrary!
Your comment focuses on the specific indicators of social status. I think that we all agree that number of followers is a very crude measure of social status in the network. But when we look at someone’s profile on a network, this is the sort of indicator that we get – not the number of shares, RTs, conversations, etc. Engagement (through replies, RTs, etc) would be a much better goal, and is the next step that many do pursue. Regarding foursquare, I do apologise for the incomplete / wrong description of how a user can acquire status within this application. I am not familiar with foursquare – I was talking about the video.
The key point of the video and the post, though, is that the forms of recognising social status are not neutral – they have implicit assumptions and they have consequences. The way the signalling systems are designed, encourage particular forms of participation. For instance, you mentioned that you could turn your house into a foursquare venue, if you want to accumulate points when you are at home. That is a very specific behaviour and way of thinking about one’s house. We need to be very conscious of these assumptions and consequences when we study the users of a particular social media platform, or incorporate that platform into our marketing programmes.
LikeLike
Hi Ana,
To be clear, my response regarding Foursquare was aimed at what was said in the video, not at you :). Besides, the leader board in Foursquare only shows the amount of points your Foursquare friends have scored, not the scores of the general 4sq population.
Anyway, it’s indeed a good point to take the type of data that a social network generates into account when assessing influence. I think that is what companies like Klout and Kred are attempting. It’s true that what you see when you look at a Twitter profile is the number of followers, and not interaction data. However, and I think this is important for academics or ‘experts’ studying this subject, the interaction data (number of RTs, listings, mentions, number of times a tweet is favorited) is often available through the API. And for those who know how to get that data through the API and then turn it into valuable information, it’s not that hard. I think that is also why it is important for those studying this subject to learn how to code. By the way: you can learn how to interact with the Twitter API on Codecademy.
My point is: if you want to study influence, you have to not only take into account that the scoring systems and generated data of social networks are not neutral and influence behaviour, but first you have to define what you mean by influence, and then dive deeper into the data sets available to see if you can get data that is closer to directly signaling influence than the superficial data of followers and leader boards. On top of that, and this is again aimed at Ms Marwick, you have to really understand the data available, and draw the right conclusions from it. In the Foursquare example she shows how not to do that. (Sure, Foursquare promotes checking in to as many places as possible, with as many people as possible; but it also rewards checking into the same venue multiple times with the same person.)
LikeLike
To be fair, the research was done between 2007-09 – So many things have changed in these applications in the meantime (including developing more sophisticated ways of measuring and showing social status).
PS – off to check codecademy 😉
LikeLike
Oh, and additionally: for marketers looking at people on specific social networks, it is just as important to really understand the data and use it properly. In the case of Twitter that could be linked to a certain hashtag (who is convincing others in a certain subject to buy stuff), and in the case of foursqare that could be tips liked (who is convincing people what to do where).
LikeLike