May 27, 2015


The EU think tanks’s tweetosphere

In my previous post on EU think tanks in the tweetosphere I noted that the Twitter community for EU think tanks is quite small, but also tightly interconnected.

A few key findings from a recent study on the political uses of Twitter, conducted around the election of the European Parliament (EP) in May 2014, are important enough to be reminded about the overall European Twitter landscape:

  • While the network architecture of the EU political tweetosphere around the EP elections is still mainly structured by language and nationality, political orientation is the most divisive factor within national clusters.
  • At the centre of the EU political tweetosphere are transnational Twitter accounts, followed by users from many different countries. Those opinion leaders on Twitter include British news media and EU organizations’ accounts, but also individual Twitter accounts set-up by journalists, bloggers, or cartoonists, which are playing significant roles -especially in national communication networks.
  • The aforementioned EU-oriented core is dominated by Western European and US-based institutional and individual accounts, with little Eastern European presence. Although at the margins, within this core, a few think tanks are also visible, namely Friends of Europe (FoE), Centre for European Policy Studies (CEPS) and Brussels European and Global Economic Laboratory (Bruegel), among Brussels-based think tanks but also OpenEurope and the Carnegie Endowment, respectively for the UK and the US.

These preliminary ‘backdrop’ findings are further confirmed by data provided by Twitter Analytics for EU Think tanks involved in the Brussels Think Tank Dialogue 2015, which will be the subject of its own post (next week). The focus on such an event is relevant as it includes the most influential Brussels-based Think Tanks of the EU political tweetosphere.

In order to allow for a quick view of the findings, the analysis is broken down in a few key questions:

  • Which are the most followed EU think tanks?
  • Are EU think tanks interacting or broadcasting in the EU tweetosphere?
  • Which are the most active EU think tanks on twitter?
  • Which are the most effective EU think tanks on twitter?
  • Which are the most trusted EU think tanks on twitter?
  • Which are the most influential EU think tanks on twitter? (Next post)

Which are the most followed EU think tanks?

The Twitter presence of EU Think Tanks is unequal in terms of followership: a minority of accounts have approximately 20,000-plus followers, a small group of accounts positioning themselves around half that audience, and the rest is hovering around the one-thousand mark (average number of followers for a Twitter account is above two hundreds).

Table 1:

Table 1

The most interesting evolution, however, concerns the decisive increase of the ratio between followers and following of all concerned think tanks, with a few exceptions highlighted in the table above. However, as showed in the same table and the figure below, there has been a steady progression of the Twitter audiences of all corporate accounts considered in the sample, with the exception of few organizations which changed their Twitter handle or opened a Brussels-specific secondary account during the considered period (Bertelsmann Stiftung and IFRI).

Figure 1:

Followership 2013 - 2015

Are EU think tanks interacting or broadcasting in the EU tweetosphere?

With the growing prominence of the EU tweetosphere, most think tanks under analysis have embraced a communication strategy anchored, at least partially, on Twitter and focused on broadcasting to their Twitter audience rather than interacting with their Twitter followers. As demonstrated by the graph below, the ratio between number of Followers and followings (F/f) of each Twitter Account is an indicator of its use of Twitter as a megaphone, as part of a broadcasting strategy on social media. In principle, the modalities of Twitter use are a good proxy of an organization’s involvement with social media.

Figure 2:

Multiples Ff

Such a broadcasting strategy is confirmed by low numbers of replies to other Twitter accounts and high number of self-referential mentions to publications, events or members of the same organisations (most of the time senior researchers and less frequently senior management).  A more conversational style is found on Twitter accounts maintained by individuals affiliated to each organisation, particularly members of the communication teams, who are re-tweeting external materials and engaging in Twitter-mediated discussions and debates. Individual accounts were excluded from a more systematic analysis, due to the original scope of this exercise.

Which are the most active EU think tanks on Twitter?

In order to gauge the presence of an overall Twitter strategy at work, it is necessary to look at a more detailed picture of Twitter activity by each organization focusing by on their output.

What can be noted from the below table is that, generally, with the exceptions of IFRI and Bertelsmann which changed their Twitter accounts during the reference period, Twitter activity has reached a sustainable level since 2013. However, in this period, a minority of think tanks invested heavily in Twitter communication, multiplying their daily output in 2 years by a factor of 8 (BRUEGEL); 6 (FRIDE); and 4 (EPC-ECIPE). The rest of the sample either maintained or increased slightly their daily output. On average, the level of Twitter communication activity across the sample was just over 4 tweets a day.

Table 2:

Table 2

It appears that the more prolific think tanks are specialising their communication teams for this task, among other things. Mainstreaming Twitter communication practices within the organisations, at the level of researchers, and particularly, of senior management, remains a challenge. It appears that social media activities –when they are managed– are handled by the communication team, with other parts of the organisation getting involved on an ‘ad hoc basis’ and/or ‘as and when’ a communication event is organised. A more systematic analysis of this aspect would require a larger enquiry including the activity of Twitter accounts of individuals affiliated to each organisation, which is beyond the scope of this exercise. [It might be worth looking at a similar study undertaken by Guerrilla Policy in the UK.]

Which are the most effective EU think tanks on twitter?

Quantity of tweets is an important parameter concerning the activity on Twitter of EU think tanks, but it says nothing about the quality of their output. Such a dimension can be evaluated by other metrics, such as the % of posts being re-tweeted as well as average number of re-tweets, as showed in the following table.

By looking at the table below, we find no general increase of third-party interest on tweets generated by EU think tanks, which remain in line with previous measurements in 2013. In fact, the dissemination effect for re-tweets is rather low across the sample, around 2 on average, with the clear exception of Bruegel which merits a separate analysis.

Table 3:

Table 3

Figure 3:

Effectivenes RTpost

In respect of 2013, there is a decrease of re-tweets per post for Bruegel, which can be partially explained by the sheer increase of its output in quantitative terms, which is confirmed by the Followers/Tweet ratio (decreasing from approximately 8 (during the year 2010) to approximately 3 currently). However, always for Bruegel, this is counterbalanced by the high number of average re-tweets. In short, the fewer tweets (from 70% to approximately 50%) that are picked up are being widely disseminated.

As a general caveat, it has to be noted that output quality, in whichever form or shape, remains an individual characteristic of the organisation, depending on off-line resources such as budget, man-power and internal quality assurance mechanisms, which as such are not the object of further investigation in this analysis.

Which are the most trusted EU think tanks on Twitter?

Another indicator for credibility and reputation within the EU-oriented tweetosphere is the number of times that each organisation is being listed in other Twitter accounts as a trusted source of information. On the one hand, listing another Twitter Account is an action less prone to be automated and is usually done by more active users of Twitter. On the other hand, listing is rather underused tool on Twitter and there are examples of misuse. Therefore, it is another metric which need to put into context, as well as total numbers of followers and tweets.

Table 4:

Table 4

In general, it can be said that, generally, EU think tanks have established their Twitter reputation in the last two years. Almost all of them are mentioned in hundreds of lists by Twitter users. Although not at the same level of leading journalists and individual bloggers, EU think tanks have and increasingly established core of followers.

In fact, it appears to be a correlation between numbers of followers and presence in Twitter lists. Both variables, however, are to be connected to the perceived reputation of the organisation on Twitter, which is connected to its influence: a much more elusive concept.

Which are the most influential EU think tanks on Twitter?

Usually, the starting point concerning a discussion of influential Twitter accounts starts with the number of followers, which provides an easy way to rank them. However, such a move equates audience with influence, which is problematic if the simple number of followers is taken at face value. In 2014, there were approximately 900 millions of open Twitter Accounts but only less than a third (255 millions) of them were active. Moreover, another 2014 estimate is that approximately 10% (one out of ten) active Twitter Accounts are fed automatically, accounting for 25% (a quarter) of Twitter traffic. For instance, in early April 2015, CEPS Twitter account proudly announced that it passed the threshold of 20,000 followers on Twitter and briefly embedded the number 20k in its own Twitter header. But a quick check of a random sample of 5000 followers, via TwitterAudit, shows that it would appear that a fifth (19%) of CEPS followers are fake.

However, if we admit that number of followers is a good starting point to discuss perceived reputation, a few precautions can be taken in order to obtain more realistic figures: for instance, verify some simple metrics for each follower, like number of tweets, date of the last tweet, and ratio of followers to friends.

Even if some EU Think Tank accounts can still boast about large numbers of ‘real’ followers, however, their quality is still more important than their quantity, meaning that EU think tanks have an interest in attracting users who are central nodes in the EU political tweetosphere. This is consistent with a two-way communication approach to social media, where a fragmented audience with scarce attention is taking its cues on EU affairs from opinion leaders, mostly journalists, bloggers and organisations, which act as informational gatekeepers.

The 2004 study on European Parliament elections sifted through 444,123 EU-related Twitter accounts and identified 11,844 Twitter users that sent at least two tweets on the subject of the European Parliament elections and had more than 250 followers. This sample was analysed using a clustering process to identify Twitter nodes shaping the information flows and discussions around EU affairs. In terms of their connections with the nodes of the EU political Twittersphere, Bruegel, Friends of Europe and CEPS appear to be more influential at the EU level rather than IFRI, SWP and Egmont, which are more present in national tweetosphere connections (respectively French, German and Belgian ones).

If being influential on Twitter means building-up a basis of followers, in either quantitative (size) or qualitative (elite) terms, this is not enough. On both accounts, a potential audience does not necessarily translate into actual engagement.

This brings us to the key question: “Does having a large and/or selected followership translate into a proportional amount of engagement?” Consequently, to understand if followership matters, we need to focus on the underlying dynamics of off-line events which are mediated on-line by Twitter to see if the audience is engaged and how.

This is the subject of my next post.

[Methodological Note:  The analysed sample was limited to 10 Brussels-based Think tanks participating in the BTTD 2015, a comparison group of 3 Think Tanks was established, composed by 3 Brussels-based Think Tanks not participating to the BTTD15, for the purpose to control for other factors and to verify the importance of participation in the BTTD.  More than 60 Data Points were collected for the main corporate Twitter feed, not including dedicated team or issue feeds, or individual staff twitter accounts, via the dedicated Twitter Analytical Platform: Twitonomy. More than 50 variables were considered, including: tweets, following, followers, listed, date of joining Twitter, ratio followers/following, ratio listed/1000 followers, tweets/day, retweets, % of retweets, user @mentions, average number of @mentions/tweet, @replies, % of @replies, links, average number of links/tweet, hashtags, average number of hashtags/tweet, tweets retweeted, proportion of tweets retweeted by others, total number of tweets retweeted, average number of tweets retweeted, users most retweeted, users most replied to, users most mentioned, hashtags most used, platforms most tweeted from.]

Read more from: Luca Barani