Changes in Twitter interactivity

The two Twitter samples allow for some comparison over how interactivity had changed over time for individuals and organisations. Firstly Graph 11. 2 looks at the changing tweeting patterns of the company Twitter accounts. In both of the official company tweets it can be seen that traditional tweets had increased from the earlier sample to the later sample. At the Leicester Mercury the informal tweets had also dropped being replaced by more formal traditional tweets. However there had been an increase in sharing tweets. Furthermore at the Bournemouth Daily Echo the interactivity had decreased over the two periods and sharing and informal tweets had remained relatively stable.

Graph 11.2: Company tweeting patterns between two samples

It would be reasonable to suggest that the decrease in interactivity and increase in traditional tweets at the Bournemouth Daily Echo was due to the same reason for the decrease in total tweets in the second sample as discussed above. The digital projects co-ordinator, part of the two person web team, was on maternity leave by the time of the second sample and she was the driver of interactivity at the case study site. This was a concern raised by the digital projects co-ordinator, Sam Shepherd (B1), to the researcher before going on leave.

What worries me about taking nine months off is that Twitter is quite a labour intensive, and if you don’t respond to people then that annoys them and if you regularly don’t respond to people then they’ll just stop asking you things and seeing we have got to the stage where it is quite a big source of information and stories I worry that that will all just disappear if they don’t get it right (B1).

However despite the web team members claiming that a “large part of the job (B1)” was responding to readers and “being interactive (B6)”, the data in Graph 11.1 shows that only about a third of their tweets were non-traditional, and this decreased over time as shown in Graph 11.2. From interviews it was apparent that the web team were predominantly using social media networks to promote the company brand, source stories and drive traffic to the website, rather than to be interactive or engage with readers. Indeed the digital projects co-ordinator kept a close eye on the number of referrals to bournemouthecho.co.uk from Facebook, stating that is was sometimes the second highest referrer after Google. Statistics cited by the web team showed that in September 2010 12,000 users were referred to the newspaper website from Facebook, compared to 2,000 in the same month the year before.

Meanwhile the changes in the @thisisleics tweeting patterns may have been due to researcher influence. During the first sample period it was also noted that the web editor responsible for administrating the @thisisleics tweets was influenced by conversations with the researcher and began to experiment with Twitter techniques such as requests to followers and more informal announcements. However this appeared to have stopped by the second sample period and reverted back to a more traditional model. This may have been also been due to a lack of organisational policy on the appropriate tweeting method from the company and a lack of leadership from the editor who preferred a laissez faire approach. It was noted in the observation period that when encouraged, the web editor tended to try new techniques but with a lack of direction she reverted back to the established gatekeeping model.

A more complex picture emerges when looking at the change in tweeting patterns by journalists and this in part may be due to how long they had been using Twitter.

Graph 11.3: Individual tweeting patterns (new users) between two samples

It is particularly interesting to note the similarities in tweeting behaviour between @martin_crowson and @StephenBaileyEcho between the two sample periods as they were both relatively new to Twitter during sample 1 as shown in Graph 11.3. Their tweeting increase or decrease was the same in all five categories. They had each decreased their informal tweets and traditional tweets and both increased their personal, sharing and interactive tweets. This would indicate that the more familiar a journalist becomes with Twitter the more likely they are to move away from a one way model to a more interactive, personal approach.

By comparison the tweeting patterns of more seasoned users @David_MacLean and @tipexxed were completely opposite in every category over the two samples as illustrated in Graph 11.4.

The account of the editor, @tipexxed, was more in line with Graph 11.3, since it indicated that a decrease in informal tweets was mirrored by an increase in personal and sharing tweets. However for @tipexxed the interactivity level was down perhaps because of more significant increases in personal (from 0% to 15%) and sharing (15% to 30%) than @martin_crowson and @StephenBaileyEcho.

Graph 11.4: Individual tweeting patterns (existing users) between two samples

The account @David_MacLean actually showed a shift back to a more traditional approach with more traditional and informal tweets and less sharing and personal tweets, although interactivity did increase. And indeed the tweets overall were still more personal and sharing than organisational accounts as shown in Graphs 1.11 and 11.2. However when looking in detail at the informal tweets it is apparent that the majority of these were comments on current affairs rather than links to the thisisleicestershire.co.uk website with a personal message or live updates and therefore were perhaps more personal in nature. The user appeared to be making more personalised comments related to his work as politics correspondent than comments about his private life outside work, whereas the other three individuals had increased their personal tweets separate from work but used informal tweets more for live updates and links to their associated website with an informal message. The different uses are discussed in more detail below.

The patterns emerging between the two individual and organisational samples appear to show that journalists are becoming more open, personal and interactive over time whilst conversely organisations are becoming more one way following the traditional model, unless they have influence from a particular individual. It may be that as the use of social media networks progresses these two clear camps will move further apart performing entirely different functions.

Talking amongst colleagues

Although journalists had higher levels of sharing and interactivity on Twitter than the official organisation accounts (@thisisleics and @Bournemouthecho) some of these tweets were to/about colleagues within the same newspaper. Therefore the content analysis considered how many of the sharing and interactive tweets were sharing colleagues’ content or sharing external users content or interacting with colleagues or interacting with external users.

Table 11.3: Division between colleague and external sharing and interaction

Username__Sharing_colleagues__Sharing_external__Interactive_colleagues'>Username	Sharing colleagues	Sharing external	Interactive colleagues	Interactive external
Bournemouthecho	24%	76%	0%	100%
martin_crowson	9%	91%	7%	93%
David_MacLean	11%	89%	0%	100%
StephenBaileyEcho	23%	76%	2%	98%
Thisisleics	24%	76%	0%	100%
Tipexxed	5%	95%	3%	97%

The table above displays the overall percentage as there was little differentiation between the two sample periods. The official organisation accounts tended to share tweets from journalists more often than journalists retweeted their colleagues messages, with the exception on @StephenBaileyEcho but overall at least three quarters of sharing was external tweets, which indicates that on Twitter newspaper organisations and individuals are sharing information from a wider network. However it could be argued that they were still acting as gatekeepers as they selected which public information to share and thus amplify. Internal interaction was even smaller indicating that journalists and organisations were using Twitter to interact with readers rather than with their colleagues.

Personal preferences

From analysing the captured tweets in more detail and comparing them with interview responses it is possible to map how individual journalists use Twitter. It has been shown in Graph 11.1 that the official Twitter accounts of newspaper organisations tend to post traditional automatic tweets, informal links to their website or share information rather than interact with readers. But individual journalists appeared to use Twitter in a variety of ways, which warrants a closer look at the four accounts analysed in this chapter.

Table 11.4: Individual journalist Twitter use

Username	Dominant coding categories	Description	Examples
martin_crowson	Equal balance between Informal and Interactive, followed by Personal	Live blogging during rugby matches and press conferences, chatting to rugby fans, talking about family life and catching up with friends	Live blogging: New Bath signing Dave Attwood being helped from the pitch by physios with what looks like ankle injury Fan interaction: @SportingbetMark It was pretty tough – as was their scrum Family life: Father’s day began at 2am with excited 4-yr-old trotting in and whispering: “How long until I can give your present Daddy?”
David_MacLean	Equal balance between Informal and Sharing, with lesser amount of Personal and Interactive	Commenting on and linking to national current affairs stories, commenting on own newspaper/blog coverage, Retweeting posts by journalists, news sources and political commentators, responding to reader comments, witty observations unrelated to work or current affairs	Commenting national: Daily Mail – “More and more families are having to withdraw their children from private education”http://t.co/L6ekgG7 - Heartbreaking Commenting on own coverage: I’ll deal with the inept spinning behind the @leicester_news Twitter account on my blog tomorrow. Retweeting: RT@GuidoFawkes I called them to ask for Hari to be stripped of his award. Responding: @mikemcsharry I’m trying to upload some PDFs to share with you all. Observations: Ordered a Gin Bramble in Cafe Bruxelles and I’ve been served something quite disgusting by mistake. It could strip paint.
StephenBaileyEcho	Mostly Informal, followed by Interactive	Commenting on local current affairs, updates/colour on own reporting activity, news/media related recommendations, responding to reader questions/comments	Updates: Just interviewed the friendliest block of pensioners ever, after a very serious flat fire on the east cliff Recommendations: The Guardian magazine had a brilliant and horrible selection of photographers best works from conflict zones today http://t.co/uiJ6gbP Responding: @GarrisonGirls Previous articles on our website, this one 95 per cent done, and should be up soon
Tipexxed	Mostly Interactive followed by Sharing and then Personal	Conversing with readers, answering reader queries, sharing Leicester lists of people to follow on Twitter, scenic photographs and posts about leisure time and family	Reader conversation: @vriyait that’s my view too, but difficult to see how he can get a fair trial what’s already been said Sharing lists: My Leicester lists is out! http://bit.ly/lhxRbq Top stories today via @leicesterpeople @housesleicester @thejamesdixon @leicesterliz Leisure time: A beautiful morning for a walk

It is significant to note that Leicester Mercury rugby correspondent Martin Crowson (L9) made a conscious decision to transfer from work related tweets only into personal tweets. In his interview he discussed how he had reached a “crossroad” and was considering carefully what step to take next:

It’s very interesting now as I have kept it very work related tweets whilst I have been doing it but you get to a stage where you start questioning whether you should do a few more things, tweet a few more things that aren’t specifically work related.........I’m at that crossroad now, where do I go with it, for that coverage because to some extent people can get certain information off the club and certain information off me and I can offer a different service, so to push that and to try and attract some more followers do I need to take that a step further and I start questioning myself, who is interested in what I’m up to? (L9)

For Martin Crowson being personal on Twitter was a way to distinguish himself from other online competition and build his own fan base and personal brand. He confirmed four months later that he had incorporated personal messages and it was definitely the right decision for him. As discussed above Martin Crowson had a high proportion of interactive and personal tweets, due to his position as a sports journalist. As he explained in his interview he spent some of his time on Twitter chatting with fans and answering questions:

If I say I have just come back from a press conference they say ‘is he fit for the weekend? Was he in good spirits? What did the head coach say today? I would like to see his reaction to X, Y, Z, let us know what he says. What do you think the score will be? Can we come back from this game?’ That sort of thing. There is good interaction (L9).

Similarly for Leicester Mercury politics correspondent, David MacLean (L12) being personal was an important part of Twitter particularly in making the most of it as a journalistic tool.

I think the personal stuff is quite important because contacts that are coming to you about, people that you would only deal with in a work environment, cabinet members at the council, they only deal with you as a journalist and I think it kind of makes you more human. What are you doing for your holidays, are you playing with your cat. It’s helped me build up a bit more of a closer relationship with people I would just usually deal with for stories.

As shown in Table 11.4 above all of the four journalists used Twitter to respond to reader questions and comments and/or have conversations with readers. The other reasons can be summarised as sharing information, commentary, live reporting and diary tweets.

Facebook comparison

As discussed in Chapter 4 the Leicester Mercury did not have a Facebook page when the first Facebook sample was taken at the Bournemouth Daily Echo. Originally the researcher planned to take two samples from the Bournemouth Daily Echo case study site to mirror the Twitter content analysis and include none from the Leicester Mercury. However after the original two sample periods the Leicester Mercury launched a Facebook page and therefore the researcher felt it was appropriate to take a sample from this to compare to the Bournemouth Daily Echo.

When analysing the findings the researcher felt it would be more helpful to analyse like with like between the two case studies and therefore decided to discount sample 1 from the Bournemouth Daily Echo and only compare the one month sample at both case study companies. Graph 11.5 is an illustration of the engagement on each of the case study site’s Facebook walls and it records how much activity there was between the company and the reader and vice versa. Each category is described in full in Chapter 5.

Graph 11.5: Facebook interaction at two case study sites

At both case study sites the largest amount of engagement was readers interacting with the content of the Facebook wall (Bournemouth Echo Sam 49%, thisisleicestershire 40%). This included posting a message on the wall or commenting on a story. Many of the comments on stories turned into conversations between readers about the original wall posting from the case study newspaper. The results also show that there was very little reciprocal engagement from the newspaper company. The Newspaper Interactive and Newspaper Sharing results were almost negligible amounting to a total of five per cent for Bournemouth Echo Sam and two per cent for thisisleicestershire. It could be argued that the Bournemouth Daily Echo Facebook account had a marginally higher Newspaper Interactive rate due to the efforts of the digital projects co-ordinator, Sam Shepherd (B1) who said: “Facebook is getting there (in terms of interaction) because people are starting to know they can Facebook chat me and I’ll answer a question and if I’m online I’ll find the answer for them.” However despite Sam Shepherd setting up a seemingly more personal profile page with the username Bournemouth Echo Sam, in contrast to the formal Bournemouth Daily Echo Facebook page, it appears that the profile page was still largely traditional in nature. The motivation for setting up the page seems again to have been to attract more readers to the brand rather than to increase interactivity as Sam Shepherd suggests: “People are more likely to be friends with an individual on Facebook rather than an organisation which might look uncool (B1)”. Therefore this profile was not a useful exemplar of how a newspaper organisation can take a more personal approach as was first thought.

The second biggest categories were Traditional (Bournemouth Echo Sam 26%, thisisleicestershire 23%) and Reader Response (Bournemouth Echo Sam 18%, thisisleicestershire 34%). The Reader Response category included when someone ‘liked’ a post or comment. It appeared once more that Bournemouth Daily Echo readers were more interactive than their Leicester Mercury counterparts, as they were more likely to comment on a post than to just simply ‘like’ a post. The opposite was true of Leicester Mercury readers who ‘liked’ posts in a higher proportion than commenting on them. This may have been due to the Leicester Mercury Facebook page being in its infancy and not having built up an interactive audience to date. It should also be noted that spam was taken into consideration when categorising the Reader Interactive engagements as postings on the wall could be promotional material. However it was found that this made up for only 11 per cent of Reader Interactive engagements at both case study sites.

It was not possible to compare how readers interacted with the case study newspaper companies on Twitter and Facebook due to their open and closed designs as explained in Chapter 4. However it was possible to compare how the newspaper companies were using the different types of social media networks to engage with their audiences. On both platforms there was limited two way interaction between journalists and readers and a traditional model prevailed, particularly on Facebook. This may have been because Twitter is a more open system which encourages interactivity whereas Facebook is a closed environment which stifles connectivity. For the two case study sites Facebook was seen as an extension of the newspaper website comment threads and a space for readers to engage with one another, rather than with the newspaper. The digital projects co-ordinator at the Bournemouth Daily Echo, Sam Shepherd (B1), explained:

What is starting to happen is people that don’t actually know each other have long conversations with each other in the comments under our stories which I think is brilliant. They are not Facebook friends but they will have a conversation and it is much more genial and less aggressive than the comments on the website and partly that will be because they are not anonymous and everyone can see who they are and so they are nicer to each other (B1).

The use of social media networks in local newspaper newsrooms does appear to be altering the traditional gatekeeping model in respect of the journalist-reader relationship. The spectrum of interactivity is shifting from a traditional approach towards a more interactive one yet the transition is still in progress. Journalists are making individual decisions to share information that would not be published in the print newspaper or on the website, and they are engaging in open conversations with their readers. However the extent to which this happens is still largely dependent on individuals rather than being incorporated into organisational norms and routines. This is in part due to organisational structures rather than journalist reluctance to embrace social media networks. This is supported by the findings of Chapter 5 and the work of Paulussen and Ugille (2008) which found that it may be difficult to create a newsroom culture of collaboration when the organisational structure does little to promote collaboration between print and online editors, IT departments and editorial and journalists and audiences. In this PhD research study the censorship of social media networks, the lack of communication between staff together with the lack of organisational leadership had created an environment where a small but growing number of journalists had decided to take the initiative themselves whilst others remain apathetic. Therefore as Heinonen (2011) suggests, integration between journalists and audiences is still limited in scope amongst mainstream news organisations. However where journalists are enabled to use social media networks there is clear indication of two way interaction and the building of relationships with readers.

The findings also illustrate that there is a clear distinction between the approach of an organisation and the approach of an individual journalist when using social media networks. Broersma and Graham (2011) identified that many news organisations use linkbots that automatically tweet headlines to generate traffic to their websites. As such they form “one-way communication” (p.2). In this PhD research there is evidence to suggest that the two case study organisations did predominantly rely on a traditional one way approach on Twitter and more strikingly on Facebook, although less traditional approaches were also being implemented. This may have been due to the smaller size of the organisations, which had thousands of readers rather than millions, meaning they had the resources to respond to readers. As explored in Chapter 2 local newspapers have more of a vested interest in connecting with their local community (Franklin and Murphy, 1998; Harrison, 1998) and therefore may be more likely to interact with their audiences than their national counterparts. That being said the findings indicate that it was individual journalists rather than their news organisation that are taking the lead in increasing interactivity with readers. As Paulussen (2011) also found individual enthusiasm and commitment are important factors for the incorporation of participatory journalism and a step by step process usually starts with a few employees before gradually spreading.

The traditional linkbot approach was negligible amongst reporters in this PhD research and instead they were increasingly engaging in a two way conversation via social media networks, an approach which Hermida (2011a) advocates, as readers increasingly gather and mediate news through their social networks online. The research discussed in this chapter suggests that two distinct channels of communication may develop in the future with a traditional function for news organisations and a social function for journalists.

As discussed above, the way in which journalists interact with readers on social media networks suggests their traditional role as gatekeepers is changing. However, although journalists use social media networks to share content, create more personal relationships and converse with their readers it could be said that these engagements serve largely traditional functions. As Hermida (2011b) suggests journalists in the main are not using social media networks to harness collective intelligence to verify information as they still prefer to rely on elite, traditional sources. Interviews with journalists at both case study sites indicated that social media networks constituted a flexible tool that could be adapted to an individual’s preferences and needs. However three main functions were indentified: increasing accountability, news gathering and building brand loyalty. Brand loyalty is one of the most significant functions and has been previously identified by Broersma and Graham (2011), Dickinson (2011) and Phillips (2011).