Information behaviour theories applied to social media: how to curate & deliver the best content

In most cases when organisations post content on their social media channels they have some agenda, something they want to achieve with that particular post. But, how can organisations monitor the success of their original agenda?

One way of measuring this success is by looking to see if it has caused our target i.e., a customer to alter their behaviour. If we succeed in getting new customers then we can have said to have changed their behaviour. If we achieve our goal of getting new customers then we succeed in changing their behaviour. This change can be said to be based primarily on information behaviour.

This is because our customer behaved according to the information that was presented to him. That is, he was interested in our product and he wanted to see where he could buy it from, and then he behaved according to how he could obtain the product or service offered. The product would be purchased and then later used.

Therefore information behaviour comes before we have the customer. But what is information behaviour?

According to Professor Wilson[1] information behavior is “the totality of human behavior in relation to sources and channels of information, including both active and passive information seeking, and information use.”

 So potential customers find themselves in a position where they reading related sources and channels of information (i.e. Facebook status, blog post, tweet about our product or service) and this can happen by active information seeking (I am looking for car insurance offer) or passive seeking (I am browsing my Facebook wall and I see a post that is of interest). Or I could use some daily offer services to look for the best offer available that day.

But then, we must ask: what can actually trigger information behavior, and what starts the process of customer information behavior? It is important to understand this so we can tailor our messages and related activities better, to achieve to convert our target into a customer. Google offered concept Zero Moment of Truth[2] back in 2012 stating that there is a challenge: to meet today’s empowered customers where they want, with the information and products they need.”

Professor Wilson’s first paper about information behavior was published back in 1981[3], which stated that:

“the aim of this paper is to attempt to reduce this confusion by devoting attention to the definition of some concepts and by proposing the basis for a theory of the motivations for information-seeking behaviour.”

So more than 30 years after Prof. Wilson’s paper was published, Google suggested that potential customers before purchasing products will need to satisfy their information needs and that companies should show up in right place, more often and with right content.

However, we still need to understand what is happening at the user side and what are the general principles of his information behavior to show up at the right place, often and with right content. To explain this we would like to reference to the recent paper “A general theory of human information behaviour .[4]This is just a brief overview of the theory and how it could be applied in the business social media activities based on it propositions. As recently Sonic Social Media analyzed leading UK food retailers we looked at their Twitter accounts and analysed their tweets through the lens of below underlined propositions.

Proposition 1. Human interaction with information results from a desire to satisfy various need states that arise in the course of human existence.

So when planning social media activities we have to address needs of those customers whom we want to change behavior and turn them into new or repeating customers.

For example Tesco’s pinned tweet states that “Don’t like spicy food but your partner does? ‘David’ makes his wife’s favorite curry – with cooling yogurt for him.[5]

We are not sure what kind of information needs (if any) this tweet addresses. This emphasizes the need to keep in mind consumers information needs when developing marketing strategies.

Proposition 2. Among those need states are problematic situations that arise in a person’s work, social relations, family life, as affected by a range of environmental factors from the socio-cultural to the physical.

Using the previous tweet from Tesco we would like to continue our theoretical analysis. Tesco’s tweet did addresses problematic situations, however it should not be pinned. Similar situations could be addressed on a continuous basis, addressing all different problematic situations based on factors from work, social relationships, and family life, as illustrated below:

Work: If you had a hard busy day order some food and bottle of wine from Tesco by replying to this tweet

Social relationships: Boost your enjoyment when watching football with your friends by ordering some chips and beer

Family Life: Surprise your mother with one month free delivery services from Tesco.

The examples could go on. There are many opportunities to address problematic situations of potential customers with tweets or Facebook statuses that are published.

Proposition 3.  The person’s motivation to seek information to satisfy the need state is affected by a range of factors, the significance of which is affected by the person’s assessment of the importance of satisfying the need state.

We can imagine a situation of a person completing a complex task and at the same time browsing through Twitter or Facebook i.e., multitasking. So factors that affect the person’s information behavior is that the person is not able to fully search through your product range.

So, why not post a message that says for example, click here and talk with our representative and find what offers are available. Because if someone is leaving work hungry and is looking for dinner the easier it is for him to see our message the more likely he could find our details of our offer, and this will increase the chances of the person buying our product. For example, one of Asda’s recent posts [6] says “Charlotte on Instagram said our Baker’s Selection Cookie Pie “tasted amazing!!!”

Yet, if someone is looking for cake, maybe they are celebrating an event, so in this context it may be better to include an offer for purchasing flowers or putting a cake in a celebration context.

Proposition 4. Having decided to seek information, the person’s ability to do so is affected by a range of intervening variables, which may be characteristics of the person themselves, or of their social relations, or of the means that exist to discover information.

So if somebody wishes to engage in additional information seeking about product or service you offer it will be affected by it social relations, his own characteristic or means on disposal to find more information. Let us look on another pinned tweet from Sainsbury’s [7] that says:

YUMMM YUMMM YUMMM Check out #fooddancing by MysDiggi x Sainsbury’s.”

So let us assume a customer likes to dance, however, what are the means on his disposal to discover more information? There is the hashtag #fooddancing but there is no link to more information about the food dancing concept Sainsbury’s is offering. There are 3.7K likes and 1.2K of shares on this particular post.

Proposition 5. Information seeking behaviour may be episodic or iterative and may be influenced by the success or failure of the actions taken.

By this proposition we have to understand that information seeking could be episodic or iterative and further digging into an offer is related to the success of previous attempts to locate information. For example Morrisons top tweet pinned on their wall states that[8] “People are raising money to build a statue of Brutus the Morrisons cat http://metro.co.uk/2017/01/27/people-are-raising-money-to-build-a-statue-of-brutus-the-morrisons-cat-in-his-favourite-spot-6409950/ … via @MetroUK”.

There is one tweet which shows interest in the results of the campaign, yet there are no replies to that tweet. Therefore, we have evidence to suggest that the user which posted this tweet will not be satisfied with the actions taken, and is less likely to engage with the company.

Proposition 6. The discovery of information may be the result of deliberate search, or accidental discovery, or information monitoring.

Therefore information which is located about our product or service could be the result of a particular search with an exact reason, unexpected discovery or information monitoring. Therefore, in posting content on social media we have to think about all three aspects: what we offer to the deliberate searcher, what we offer to the Twitter user who visits out post accidently, and finally to those users who monitor our posts regularly.

Looking at the tweets of Morrison we discussed above, the user who comes to this account accidentally would not see the company values and would hardly associate with them even if he had the same values as a company. Therefore it is important to always stream the content that satisfies all three different types (deliberate search, accidental discovery and information monitoring) of the user that could visit our social media channel.

Proposition 7. Information seeking is only one aspect of information behaviour: other activities (which may play a part in information discovery) include information exchange or sharing, information transfer to others whose needs are known, as well as the avoidance and rejection of information.

So we have to be aware that not all visitors of our social media channel are actually seeking for information, but they could come there as a response if someone shared content with them. Also some users would like to avoid content we are distributing. This proposition is challenging as it includes two extremes, avoid information (somebody is not interested at all in the content we provide) and share information (that means somebody loves the content we provide).

However, if we are aware of these extremes then we can adapt our content production strategy with this information in mind. Looking back at Sainsbury’s tweet: YUMMM YUMMM YUMMM Check out #fooddancing by MysDiggi x Sainsbury’s.” we can imagine that some Twitter users will love the idea of food dancing and are likely to share the post further, however the Twitter users who do not like to dance would avoid the tweet.

In this instance, it would be better to offer the same post options for Twitter users who would not like to dance. For instance, providing a link to the receipt i.e., if you do not feel dancing check our receipt of that meal they are dancing to,

Proposition 8. Information behaviour may be individual, collective or collaborative.

This proposition could be very important in development of the content, as we have to assume that we are addressing individuals, collective and collaborative information seeking targets. Most of the content produced for marketing purposes of social media are usually aimed towards individual information behavior. For example retail chains we address in this text could balance with their tweets and address in balance individual, collective and collaborative information seeking processes. Therefore, organisations could offer the ability to ‘share the bill’, so as to allow potential customers to participate in sharing the costs of the party ingredients together.

The ideas and concepts mentioned above are not exclusive, but are used to give examples of how the general theory of information behaviour can be applied in social media marketing activities. The post aims to demonstrate that advance machine learning activities are not required in developing superb social media tactics. We simply need to be aware of what can trigger the information behaviours of potential customers, and that we should have those propositions mentioned above close to hand when developing social media content strategies.

[1] Wilson, T. D. (2000). Human information behavior. Informing science3(2), 49-56.

[2] https://www.thinkwithgoogle.com/research-studies/2012-zmot-handbook.html

[3] Wilson, T. D. (1981). On user studies and information needs. Journal of documentation37(1), 3-15.

[4] Wilson, T. D. (2016). A general theory of human information behaviour. Information Research21(4). http://www.informationr.net/ir/21-4/isic/isic1601.html

[5] https://twitter.com/Tesco/status/818366973708890112 retrieved 27.01.2017

[6] https://twitter.com/asda/status/825063351021481984 retrieved 27.01.2017

[7] https://twitter.com/sainsburys/status/821507407146663936 retrieved 27.01.2017

[8] https://twitter.com/Morrisons/status/825057872270807040 retrived 27.01.2017

A post by Sergej Lugovic and edited by Wasim Ahmed.

Comparing Tesco, Asda, Sainsburys, and Morrisons on Twitter: who wins?

 

Below is a comparison using very powerful analysis tool NodeXL which takes a look at and compares the Twitter networks of Tesco, Asda, Sainsburys, and Morrisons.

Twitter is a social media platform, and also a news monitoring platform, it allows you to monitor not only your own performance, but also your competitors.

NodeXL network graphs cluster all mentions of your organisation into a number of different groups, and by doing so patterns emerge in the form of different groups. Associated with the patterns are metrics such as:

  • Top URLS
  • Top KEYWORDS
  • Top HASHTAGS
  • SENTIMENT
  • TOP WORDS
  • TOP CO-WORDS
  • TOP INFLUENCERS 

The figure below provides a 4 way comparison between the different supermarket chains, and instantly we can note some similarities and differences:

  • Each supermarket chain has a large broadcast network where users are retweeting and engaging with tweets and/or Twitter account of a supermarket
  •  All Twitter accounts have an isolates group and some are larger than others, for instance, it appears Asda have the largest group of isolates
  • A low graph density (number of users connected to each other) may be describable and  Morrisons has the lowest followed by Tesco, Sainsburys, and Asda
  • Sainsburys has the most users, and URLs NOT connected to its brand tweeting about the supermarket, for instance it has mentions from the Guardian
  • Tesco and Morrisons have a fair level of internal and external mentions
  • Asda has the most users and URLs from within the brand tweeting and sharing information related to the supermarket
  • The hashtag VEGAN appears in Tesco and Sainsburys network graphs
  • Tesco has the largest user-base on Twitter with a total of 17,621 unique users over the last week or so, followed by: Sainsburys (15,087), Asda (11,117), and Morrisons (1658)

4 way comparison =

Comparision.png

Using the metrics above we can rank the supermarkets in the following order:

  1. TESCO

  2. Sainsburys

  3. Asda

  4. Morrisons

 

Now lets take a closer look at each of the network graphs:

1. Tesco

TESCO.png

 

READ OUR FULL REPORT ON TESCO HERE

In addition the the network visualization above, a number of metrics are also produced alongside the graph. These metrics are produced OVERALL, and by GROUP-LEVEL.

There are many, many metrics produced, and although we provide a sample of these below we recommend that you click the link above and explore the network graph and metrics further. 

In Tesco’s graph one of the most frequently occurring URLS is that of the news that Tesco has been voted as the most improved supermarket in the UK. The top hashtags are that of tesco, deal, and everylittle helps among others. 

Top Hashtags in Tweet in Entire Graph:
[239] tesco
[56] deal
[47] everylittlehelps
[45] essothursdays
[40] waspi
[36] vegan
[35] blackfriday
[30] cheapgames
[24] poorservice
[18] happytohelp

Top Words in Tweet in Entire Graph:

[12682] tesco
[5017] back
[4842] 10
[4834] card
[4831] aavaiz

Top URLs in Tweet in Entire Graph:

[6682] http://www.jklj.asia
[487] http://directlink.jp/tracking/af/1442580/93NGPKX7-JTt17LSM/
[133] http://du3a.org
[59] http://www.jagran.com/punjab/tarantaran-sadas-asda-15417426.html
[54] https://www.amazon.co.uk/Year-Being-Single-Fiona-Collins/dp/0008211469/ref=tmm_pap_swatch_0?_encoding=UTF8&qid=&sr=
[51] https://www.g2a.com/r/cheapfm2015equityxd
[50] https://twitter.com/kylndaclouds/status/790655392615399428
[26] https://www.retailgazette.co.uk/blog/2017/01/asda-co-op-the-only-retailers-listed-on-top-lgbt-inclusive-list
[24] http://asda-grocery.custhelp.com/app/answers/detail_grow/a_id/1738
[21] http://www.talkandroid.com/307114-sponsored-asda-mobile-sim-only-deals-holiday-2016/

 

2. Sainsburys

SAINSBURYS.png

READ OUR FULL REPORT ON SAINSBURYS HERE

In addition the the network visualization above, a number of metrics are also produced alongside the graph. These metrics are produced OVERALL, and by GROUP-LEVEL.

There are many, many metrics produced, and although we provide a sample of these below we recommend that you click the link above and explore the network graph and metrics further. 

An article to appear among the most frequently occurring URLs within the network graph was that of an article published by the Guardian titled Campaigners call on EU to halve food waste by 2030. The most frequently occuring hashtags consisted of fooddancing, newyearsresolution and vegan among others. 

Top Hashtags in Tweet in Entire Graph:
[1739] fooddancing
[412] newyearsresolution
[331] sainsburys
[327] win
[137] vegan
[137] foodwaste
[109] dryjanuary
[108] wastelesssavemore
[89] courgettecrisis
[79] tuesnews

Top Words in Tweet in Entire Graph:

[11323] sainsburys
[3115] yummm
[1968] sainsbury’s
[1950] out
[1751] fooddancing

Top URLs in Tweet in Entire Graph:
[426] https://twitter.com/sainsburys/status/821507407146663936
[293] https://www.theguardian.com/environment/2017/jan/18/eu-proposals-halve-food-waste-europe-2030-uk-supermarkets-tesco-sainsburys
[186] https://twitter.com/login?redirect_after_login=%2Fmessages%2Fcompose%3Frecipient_id%3D80685646
[180] https://www.youtube.com/watch?v=bUphaBWcCrw
[167] https://www.theguardian.com/environment/2017/jan/18/eu-proposals-halve-food-waste-europe-2030-uk-supermarkets-tesco-sainsburys?CMP=share_btn_tw
[104] https://www.theguardian.com/lifeandstyle/2017/jan/22/all-change-supermarket-aisles-more-veg-sainsburys-cut-meat-consumption?CMP=share_btn_tw
[104] https://wastelesssavemore.sainsburys.co.uk/saving-tips/10-mocktails-you-can-make-right-now?adb_src=2-b56-c3_21.01.2017
[87] http://www.express.co.uk/life-style/science-technology/754302/Sainsburys-250-gift-voucher-beware-WhatsApp-scam
[80] https://www.theguardian.com/environment/2017/jan/18/eu-proposals-halve-food-waste-europe-2030-uk-supermarkets-tesco-sainsburys?CMP=twt_gu
[72] https://www.eveningtelegraph.co.uk/fp/ex-cop-slams-officers-dundee-sainsburys-armed/

3. Asda

ASDA.png

READ OUR FULL REPORT ON ASDA HERE

In addition the the network visualization above, a number of metrics are also produced alongside the graph. These metrics are produced OVERALL, and by GROUP-LEVEL.

There are many, many metrics produced, and although we provide a sample of these below we recommend that you click the link above and explore the network graph and metrics further. 

A URL to appear among the most frequently occurring list for Asda’s network graph is that of the news that Asda and Co-op were the only retailers to be listed on top LGBT-inclusive list. The most frequently occurring hashtags consist of deal, gggthefil, and black friday among others.

Top Hashtags in Tweet in Entire Graph:

[109] deal
[108] gggthefilm
[84] furniture
[57] blackfriday
[49] productoftheyear
[44] cheapgames
[41] goodliving
[41] ophercules
[37] fridayfeeling

Top Words in Tweet in Entire Graph:

[9449] asda
[3209] gt
[3125] lt
[741] tesco
[637] 速報

Top URLs in Tweet in Entire Graph:
[6682] http://www.jklj.asia
[487] http://directlink.jp/tracking/af/1442580/93NGPKX7-JTt17LSM/
[133] http://du3a.org
[59] http://www.jagran.com/punjab/tarantaran-sadas-asda-15417426.html
[54] https://www.amazon.co.uk/Year-Being-Single-Fiona-Collins/dp/0008211469/ref=tmm_pap_swatch_0?_encoding=UTF8&qid=&sr=
[51] https://www.g2a.com/r/cheapfm2015equityxd
[50] https://twitter.com/kylndaclouds/status/790655392615399428
[26] https://www.retailgazette.co.uk/blog/2017/01/asda-co-op-the-only-retailers-listed-on-top-lgbt-inclusive-list
[24] http://asda-grocery.custhelp.com/app/answers/detail_grow/a_id/1738
[21] http://www.talkandroid.com/307114-sponsored-asda-mobile-sim-only-deals-holiday-2016/

4. Morrisons 

MORRISONS.png

READ OUR FULL REPORT ON MORRISONS HERE

In addition the the network visualization above, a number of metrics are also produced alongside the graph. These metrics are produced OVERALL, and by GROUP-LEVEL.

There are many, many metrics produced, and although we provide a sample of these below we recommend that you click the link above and explore the network graph and metrics further. 

A URL that appeared within the graphs most frequently occurring list is that of an article titled: The very strange reason that drivers keep pulling over for Morrisons delivery vans, which notes how emergency workers are mistaking Morrison vans for ambulances. The most frequently occurring hashtags  are that of fintech, morrisons, and uk among others. 

Top Hashtags in Tweet in Entire Graph:
[45] fintech
[27] morrisons
[26] uk
[24] auspol
[18] voteclicsargent
[14] london
[14] eu
[14] scomo
[14] deal
[13] bitcoin

Top Words in Tweet in Entire Graph:

[1078] morrisons
[278] delivery
[240] van
[239] morrisons’
[235] ambulance

Top URLs in Tweet in Entire Graph:

[27] http://www.afr.com/news/politics/scott-morrisons-uk-fintech-mission-20170122-gtw7fz
[20] http://www.manchestereveningnews.co.uk/news/greater-manchester-news/very-strange-reason-drivers-keep-12490731
[16] https://www.thesun.co.uk/news/2670823/paramedic-accidentally-mistakes-a-morrisons-delivery-van-ambulance/
[15] http://www.laweekly.com/music/searching-for-jim-morrisons-ghost-in-gentrified-venice-7830641
[12] http://www.afr.com/news/politics/scott-morrisons-uk-fintech-mission-20170122-gtw7fz?utm_content=buffer0f36f&utm_medium=social&utm_source=twitter.com&utm_campaign=buffer
[11] http://www.theladbible.com/more/uk-morrisons-vans-so-much-like-ambulances-even-paramedics-get-confused-20170122
[10] http://www.thepoke.co.uk/2017/01/17/dough-morrisons-plays-ball-cock-ring-donut-customer-complaint/
[8] http://www.dailytelegraph.com.au/remote/check_cookie.html?url=http%3a%2f%2fwww.dailytelegraph.com.au%2fnews%2fnational%2fscott-morrisons-trade-deal-plan-for-brexit%2fnews-story%2f8341d7f6722a32ab404ff8039ff4fad3
[7] https://twitter.com/MikkiL/status/823217653296807936
[6] https://www.g2a.com/r/cheapfm2015equityxd

FULL REPORTS:

READ OUR FULL REPORT ON TESCO HERE

READ OUR FULL REPORT ON SAINSBURYS HERE

READ OUR FULL REPORT ON ASDA HERE

READ OUR FULL REPORT ON MORRISONS HERE

What do Russell Group Universities look like on Twitter?

In our January give-away we used The Complete University Guide to locate universities which form part of the Russell Group.  We then took a sample of Universities from the list in order to run a comparison.

The universities that we selected consisted of:

  • Oxford University
  • Nottingham University
  • Durham University
  • Manchester University
  • Leeds University

compar

Using powerful network metrics developed and perfected over a number of years we were able to visualize the Twitter accounts of these Universities. The graphs look more complicated than they actually are (we promise!)

With a bit of guidance from one of our team members you too can become an expert in reading these graphs. For each graph we have provided a link to the full report, which contains things like popular URLs, keywords, hashtags, domains, and words.

If you are associated with the Digital Team at a UK higher education institution, we are able to provide (a free) non-obligation overview on interpreting these graphs.

An observation among the network graphs is that each university Twitter account has quite a large broadcast network as the largest group of Twitter users. This is on the left hand side of the network graph, labeled G1.

Broadcast networks occur when a user is retweeted with high frequency, for instance, news organisations and authors tend to have quite large broadcast networks.

University of Oxford  

oxford.png

The most frequently occurring words in the largest broadcast network consist of: uniofoxford, inequality, oxford, years, tony, university, work, here, read, more.

The most frequently occurring URL in group 1 is:

[68] http://hpc-asia.com/morten-middelfart-big-data-solutions-for-tumor-sequencing

The full report can be accessed here

Nottingham University 

Nottingham.png

The most frequently occurring words in the largest broadcast network consist of: uniofnottingham, antibiotic, new, year, uni, 2017, spider, meeting, silk, spider, silk.

The most frequently occurring URL in group 1 is:

[32] http://phys.org/news402738616.html

The full report can be accessed here.

University of Durham  

DURHAM.png

The most frequently occurring words in the largest broadcast network consist of: durham_uni students, congratulations, durham, graduating, today ducongregation, day, conference, student. The full report can be accessed here.

The most frequently occurring URL in group 1 is:

[6] http://www.thetimes.co.uk/edition/news/student-jailed-for-sex-assault-on-sleeping-victim-

Manchester University

officialuom

The most frequently occurring words in the largest broadcast network consist of: officaluom, thanks, here, photo, instagram, campus,year, new, out, find. The report can be accessed here.

The most frequently occurring URL in group 1 is:

[10] http://www.manchester.ac.uk/discover/news/clues-from-past-volcanic-explosion-help-manchester-led-team-model-future-activity/

Leeds University  

leeds.png

The most frequently occurring words in the largest broadcast network consist of: univeristyleeds, new, students, death, help, post, rasberry_pi, build, star, kinda. The full report can be accessed here.

The most frequently occurring URL in group 1 is:

[12] http://www.wired.co.uk/article/jet-stream-earth-core

How can we use this information?

Social media managers, and marketing professionals can utilize these network graphs for intelligence, such as:

  • Popular content generated on a weekly, daily, or monthly basis
  • Ability to monitor content on rival universities

Our pricing is very competitive, and we are confident that we can provide the best possible deal.  It is important to note that the graphs are topical and vary by time. The graphs can be created on a weekly, and/or monthly basis.

What does @ddjournalism look like on Twitter?

In this post we take a look  firstly at the @ddjjournalism Twitter account, and then the #ddj hashtag. Data Driven Journalism (ddj) is a fantastic resource, and has a number of excellent articles related to that of social media. It is a hub for news and resources from the community of journalists, editors, designers, and developers who use data in the service of journalism. Check out an article on NodeXL that was published on the ddj which can b found here.

Figure 1 – Network graph of @ddjjournalism (see full report here)ddj.png

The visualization looks great, but you may be asking, what does this all mean?

The top left hand side of each group is labeled e.g. G1 refers to ‘Group 1’ and G2 refers to ‘Group 2’ and so forth, and the keywords relate to the most frequently occurring per each group. NodeXL network graph reports also produce a number of other metrics such as the most frequently shared URLs, Domains, Hashtags, Words, Word Pairs, Replied-To, Mentioned Users, and most frequent tweeters These metrics are produced overall and also by group of Twitter users. By looking at different metrics associated with different groups (G1, G2, G3 etc) you can see the different topics that users may be talking about.

We can also take a look at the 6 types of network structures from  Smith, Rainie, Shneiderman, & Himelboim (2014) as a guide in interpretation.

Figure 2 ‘Six Types of Network Structures’

12345

Looking back at figure 1, we can now see that many of the groups resemble a broadcast network. That is, users are sharing articles that have been published on the website, and their followers are retweeting these.

Next we decided to take a look at the hashtag ‘#ddj’ to examine some of the top influences and popular content associated with data driven journalism.

Figure 3 Network graph of ddj (full report here)

ddj dataviz.png

We can then take a look at some of the top influences (ranked by betweenness centrality):

Figure 4 – Top influencers ranked by betweenness centrality 

betweenness-centrality

The figure above displays a number of influencers, ranked by betweenness centrality, related to the hashtag ‘#ddj’, which includes YouTube, the Financial Times, and individuals sharing ddj related news.

Figure 5- Most frequently occurring URLs

top-urls

The data driven journalism resource may be interested to know that their survey which sought information on the current status of data journalism was among the most popular URLs that were shared within the previous week or so.

We used Twitonomy to chart some analytics to look at the most frequent days of the week and hours of the day that the @ddjjournalism is most active.

Figure 6- Days and hours of the week most active 

days-and-hours-123

As the figure above shows the most active day for the @ddjjournalism account is that of Thursday, and the most popular time for sharing content is 10am.

#WeAreInternational campaign demonstrates UK remains diverse & inclusive to international students & staff

The #WeAreInternational initiative traces its roots  to 2013 whereby those at the University of Sheffield wanted to eradicate the myth that international students did not integrate with the city.

It has so far has been backed by over a 100 universities, education institutions, government organisations as well as international organisations. The full list of supporting organisations can be found here.

In this blog post we use a very powerful brand monitoring tool Visibrain to gain insight into the impact that the campaign has been having.

Figure 1 – Time series graph which shows when there were peaks in tweets

time-series-graph-weareinternational

The figure above, created using Visibrain, displays the peaks in the hashtag over a 30 day period. The first peak corresponds to the 25th of October and this peak occurred as a number of organisations and individuals tweeted their support of the hashtag on this day. These included the Foreign Office, the editor of the World University Rankings, and Cardiff University among others. The largest peak of tweets occurred on the 9th of November, and correlated to a #WeAreInternational event, which took place with the University of Sheffield’s Vice Chancellor, in Delhi.

Looking carefully at some of the numbers generated using Visibrain from October 22nd to November 21st the initiative had a total of 8,782 tweets which were tweeted out by 4,426 users, and which had 36,804,730 impressions, with 2,061 (23%) original tweets, and 6,721 (77%) retweets, and 3,289 (37%) tweets contained links. In interpreting these results, firstly this indicates that the hashtag has quite a high retweet percentage, and a possible reason for this is that users wish to indicate their support of the initiative, and retweet in order to have the hashtag appear on their timelines. The number of impressions (a metric Twitter users to refer to views) is also quite high which may indicate that users with large followings may have been tweeting and retweeting the hashtag. Next we took a look at the most frequently occurring words, as shown in figure 2 below.

Figure 2 – Most frequently occurring words

word-cloud

The most frequently occurring word is that of #WeAreInternational (n=7951), followed by #LoveInternational (n=391), GlobalBritain (n=335), #Brexit (n=230), and #WorldWeek16. The aim of the campaign is to highlight how the UK remains diverse and inclusive to international students, and many of the words used, such as ‘#globalbritain’ suggest this. We can also see that a number of countries are also mentioned such as Malaysia, Nambia, China, and India. Next we take a look at where the users tweeting come from.

Figure 3 – Top 10 locations of user tweeters

countries

In the top ten table  the majority of tweets derive from the United Kingdom (81.2%). However, those behind the initiative would be pleased to see that 18.8% of tweets derive from outside of the UK. This would suggest that the hashtag is not only reaching those outside of the UK, but that it is also being engaged with via original tweets, and retweets. Next we used NodeXL and took a look at the structure of the network for tweets which were tweeted over the time period of 25 November 2016 to 04 December.

Figure 4 – NodeXL Network graph (access full report here)

weareinternation123

NodeXL network graphs can take a number of different shapes, and this post may be useful for those looking at this for the fist time. The graph above has clustered Twitter users into a number of different groups, and many of the groups represent broadcast networks (G1 to G8). In a broadcast network a user may send out a tweet and their followers or a group of users following a topic may retweet the user. The complete NodeXL report can be found here.

This initiative demonstrates that social media can be used to raise awareness and disseminate information on an unprecedented scale. This blog post referred to data and analytics powered by Visibrain, and Connected Action

What does @marshawright look like on Twitter?

Marsha Wright is a well distinguished business expert, an entrepreneur, and is easily one of the most successful Twitter users the platform hosts. Users may be familiar with the  initiative, created by Marsha, which is a trending movement which has many millions of shares, and has been featured as a trending topic on Twitter hundreds of times.

In this post we used NodeXL to generate a network graph and associated report based on the Twitter handle ‘@marshawright‘. The output can be found below in figure 1 and displays Twitter activity from the previous week or so.

Figure 1 – Network graph of ‘@Marshawright’

wasim1234

The visualization looks great, but you may be asking, what does this all mean?

The top left hand side of each group is labeled e.g. G1 refers to ‘Group 1’ and G2 refers to ‘Group 2’ and so forth, and the keywords relate to the most frequently occurring per each group. NodeXL network graph reports also produce a number of other metrics such as the most frequently shared URLs, Domains, Hashtags, Words, Word Pairs, Replied-To, Mentioned Users, and most frequent tweeters These metrics are produced overall and also by group of Twitter users. By looking at different metrics associated with different groups (G1, G2, G3 etc) you can see the different topics that users may be talking about.

We can also take a look at the 6 types of network structures from  Smith, Rainie, Shneiderman, & Himelboim (2014) as a guide in interpretation.

Figure 2 ‘Six Types of Network Structures’

12345

As show in the figure above different topics on social media can have contrasting network patterns. For instance in the polarized crowd discussion one set of users may talk about Donald Trump and others about Hilary Clinton, in the unified crowd users may talk about different aspects of the election, and in brand clusters people may offer an opinion related to the election without being connected to one another and without mentioning each other.

In a community cluster a group of users may talk about the different news articles surrounding Hilary Clinton. Broadcast networks are typically found when analysing news accounts as these disseminate news which is retweeted by a large amount of users. We can think of support networks as those accounts which reply to a large number of accounts, we can think of the customer support of a bank which may reply to a large amount of Twitter users.

Now if we take a look at Marsha’s graph again:

Figure 1 (revisited) – Network graph of ‘@Marshawright’

wasim1234

We can see that G1 (Group 1) resembles a broadcast network where @Marshawright’s account is in the centre of the group and accounts around the group are those which are retweeting and/or engaging with Marsha’s account. This is quite an impressive broadcast network and resembles that of news media organisations. We can also see a number of smaller groups which also resemble a broadcast network, which may indicate that Marsha’s tweets tend to have a high level of share-ability.

The full report can be found here, and there are quite a lot of analytics produced alongside the network graph. In our blog posts we like to highlight different metrics that NodeXL can produce. In this post we take a look at words overall and the level of sentiment which is produced based on tweet content.

Top Words in Tweet in Entire Graph:

[4252] Words in Sentiment List#1: Positive

[1071] Words in Sentiment List#2: Negative

[27] Words in Sentiment List#3: Angry/Violent

[60120] Non-categorized Words

[65445] Total Words

[2991] thinkbigsundaywithmarsha

[1819] marshawright

[548] https

[432] quotes

[402] entrepreneur

The information above, not surprisingly, displays that the most frequently occurring word in Marsha’s network graph is ‘ thinkbigsundaywithmarsha’ (n=2911). This is a topic that is tweeted about by followers of the initiative on a weekly basis and which involves sharing positive quotes by Twitter users. The graph also displays the sentiment of the words in discussions that involve Marsha, and we can see here clearly, that discussions tend to be positive as there number of words in the positive sentiment list outnumber those in the negative sentiment list.

The graphs are topical and are likely to alter on a week by week basis, however, it is very likely to be the case that the keyword ‘thinkbigsundaywithmarsha‘ would appear in future NodeXL graphs and reports. To further highlight this point we can examine a time series graph created in Visibrain, which charts the Twitter activity of ‘thinkbigsundaywithmarsha‘, including keywords and hashtags.

thinkbig

We can see that every-weekend (when followers of Marsha are encouraged to tweet using the hashtag) there is a quite a large peak in tweets. Therefore, we can conclude that this is keyword which will be present in future network graphs and reports.

Here at Sonic Social Media we like Twitter influencers, and so we offer completely free of charge analysis of accounts which belong to those whom identify as Twitter influencers. You can complete this form and we can have something back to you within 48 hours.

Want to learn more? Check out this NodeXL and Social Network Analysis workshop at Leeds Becket University which will cover all of the methods used in this post in much depth.

What does @PeterJukes look like on Twitter?

In this blog post we provide some insight into the the Twitter account of @PeterJukes, a journalist,  English author, screenwriter, playwright, literary critic and blogger. We follow a similar methodology to that of previous work where analysed the Twitter account of Paul Mason. 

Peter, is a prolific Twitter user with 199 thousand tweets, and 23 thousand followers. So we thought that his account would make for some good analysis. We used the Fruchterman Reingold graph layout algorithm produced using NodeXL which displays the Twitter network of Peter Jukes. The tiny boxes represent nodes and this case these are Twitter users, and the lines represent ‘edges’ which are the lines connecting Twitter users.

The network graph below displays all the Twitter accounts that have engaged with the Twitter handle ‘@PeterJukes‘  over the previous seven or so days.

Figure 1 – Network graph 1 using the Fruchterman Reingold graph layout algorithm

Peter Jukes.png

The image above displays the Twitter network of the ‘@PeterJukes’ Twitter handle. See the NodeXL graph gallery for the full image, alongside a detailed report, and an interactive explorer.

We can use the interactive explorer functionality to zoom into the network:

Figure 2 – Using the interactive explorer to zoom into network of ‘@PeterJukes’

peter jukes2.png

In the figure above we have hovered over the ‘@PeterJukes‘ Twitter handle which is towards the centre of the network, and the accounts closest to that of Peter’s account indicate strong connections i.e., they may be accounts that are frequently mentioned or those which engage with Peter’st tweets the most. The connections between Twitter users are directed, for instance, Donald Trump appears in the graph as he has been tagged in tweets, similarly to that of CNN.

This is one way of laying out Twitter networks, but one of our favorite methods is by using NodeXL to select the Harel-Koren Fast Multiscale algorithm to layout the graphs.  This finds groups and divisions within a Twitter network, as it clusters Twitter topics based on the same network into different groups. Discussions on Twitter can form into a number of different groups, see figure 2 below for 6 types of networks that can form.

Figure 3 – Six Types of Twitter Network (see  Smith, Rainie, Shneiderman, & Himelboim, 2014)

12345

As show in the figure above different topics on social media can have contrasting network patterns. For instance in the polarized crowd discussion one set of users may talk about Donald Trump and others about Hilary Clinton, in the unified crowd users may talk about different aspects of the election, and in brand clusters people may offer an opinion related to the election without being connected to one another and without mentioning each other.

In a community cluster a group of users may talk about the different news articles surrounding Hilary Clinton. Broadcast networks are typically found when analysing news accounts as these disseminate news which is retweeted by a large amount of users. We can think of support networks as those accounts which reply to a large number of accounts, we can think of the customer support of a bank which may reply to a large amount of Twitter users.

Figure 4 – Network graph 2 using the  Harel-Koren Fast Multiscale algorithmunclustered.png

The network graph is made up of several groups of Twitter users. Group 1 displays the Twitter audience of @PeterJukes, which is has elements of Broadcast Network which indicates users are retweeting Peter, but also that of a Support Network, as Peter is replying to and engaging with other Twitter users.

The top left hand side of each group is labeled e.g. G1 refers to ‘Group 1’ and G2 refers to ‘Group 2’ and so forth, and the keywords relate to the most frequently occurring per each group. NodeXL network graph reports also produce a number of other metrics such as the most frequently shared URLs, Domains, Hashtags, Words, Word Pairs, Replied-To, Mentioned Users, and most frequent tweeters These metrics are produced overall and also by group of Twitter users. By looking at different metrics associated with different groups (G1, G2, G3 etc) you can see the different topics that users may be talking about.

The full report can be accessed here, and there are enough analytics in the report to write an entire PhD thesis, but for the purposes of brevity, in this post we can examine some of the most frequently occurring co-words (words that appear together), keywords, and hashtags, and articles to get a sense of what types of topics have dominated Peter’s network over the last week.

Top 4 Hashtags  in Entire Graph:
1. [180] trump
2. [93] brexit
3. [17] peterjukes
4. [14] pmqs

Top 4 Word Pairs in Entire Graph:
1. [169] piece,read
2. [168] read,obama’s
3. [168] obama’s,attitude
4. [168] attitude,trump

Top 4 URLs
1.[49] A link to Peter’s tweet which links to a New York Times article related to Obama and Trump
2.[44] A link to a Guardian article titled ‘All Out War; The Brexit Club; The Bad Boys of Brexit review – rollicking referendum recollections’
3.[15] A link to a Who What Why article titled ‘ A Dark View From Flyover Country ‘ 
4.[11] A link to a Politics.co.uk article titled ‘The British press is an enemy of free speech’

Examining the network graph and the most frequently occurring hashtags, URLs, and words, we can conclude from the data that a popular topic among the network over the past week was related to that of article related to Obamas view’s towards Donald Trump, and specifically this tweet which was sent by Peter promoting the article. It is important to note that these graphs are topical and change on a week by week basis, we would recommended browsing the full report on the NodeXL graph gallery.

Just for a bit of fun, and because we are interested to examine the days of the week and times journalists tend to tweet. We used Twitonomy, to examine the average time and day Peter tweets, as shown in figure 5 below.

Figure 5 – Days of the week and time tweeted

what-days-active

It looks as though Saturday is the most active day, with 9pm as the time of choice for sending tweets.

Here at Sonic Social Media we like journalists, and so we offer completely free of charge analysis of accounts which belong to journalists. You can complete this form and we can have something back to you within 24 hours.

Want to learn more? Check out this NodeXL and Social Network Analysis workshop at Leeds Becket University which will cover all of the methods used in this post in much depth.

Why use a network approach for social media monitoring?

Social media managers or digital teams may be faced with a number of questions such as:

  • How do we increase the visibility of our messages?
  • How can we increase the number of followers, likes and retweets?
  • How do we become top influencers around certain discussions?
  • How can we make some of our messages viral?
  • How do we gain actionable insight?

From a network point-of-view this translates to:

  • How do we build a network reach?
  • What divisions or groups are present when users mention our brand?
  • Who are the most influential people in the discussion?
  • What exactly are they talking about?

A key benefit of social media network maps and reports created with NodeXL is to bypass the need to read thousands of tweets and messages on a range of topics.

NodeXL reports  can be used for measuring and monitoring not only your own, but also your competitors´ performance.

At the highest level, a network approach allows social media managers to recognize that the shape of their crowd is different from the optimal shape, and use network metrics to guide the transition between the current and desired state.

What is the structure of your brand? Is the structure of your network polarized? Or is it a brand cluster? Figure 3 from Smith, Rainie, Shneiderman, & Himelboim, 2014 provides a guide in contrasting patterns within network graphs:

Figure-3

So, what does the structure of your brand look like on social media?

Get in touch for a consultation on how you can better understand the discussion on your brand, identify key performance indicators , and how you too can gain actionable insight. See maps that have been created by the Connected Action Team here.