Author: 
Catriona Bryce
Publication date: 
Tuesday, 23 February, 2016
Abstract: 

Using social media is now business as usual for the National Library of Australia and Trove. The Trove team believes that it is engaging people in these spaces, people who know and love Trove and those who have never heard of it. The team believes it is promoting the use of Trove. But is this true? Do people really click through to those links? Do people find Trove and keep coming back? Understanding what the team at Trove wants to achieve in social media is crucial to assessing its value. Is impact in the numbers, the individual stories of lives changed, or is it an interweaving of both?

It all starts with a poodle

Trove’s most ‘successful’ moment in social media involved a poodle. Cachi the poodle fell out the window of an apartment building in Buenos Aires in 1988 (The Canberra Times 24 October 1988, p.1). As a result of his fall, three people died. This article, found in Trove’s digitised newspapers, was posted on the social sharing site Reddit. Before staff knew it, Trove experienced its single biggest usage surge. Over the course of three days, there were 234,644 visits to this article in Trove. The Trove team did not share it. Staff did not even know it existed until the traffic to Trove was seen and investigated, but it led thousands of people to Trove. Did this encourage people to stay and read more? Did they return? The team's reading of Google Analytics suggests that, for the most part, it did not and they did not. It was an anomaly: an interesting one, but not one that achieved the obvious aims for a presence in social media, which is to inspire sustained increased usage. It did not increase engagement. It did not grow a community. There were many clicks over the course of a couple of days, but that is all.

During the 2015 ALIA Online conference I followed the conference hashtag to capture all mentions of Trove. While monitoring this feed I noticed some references to Douglas Adam’s A Hitchhikers Guide to the Galaxy, a seminal work for librarians with a certain sense of humour. While tweeting as Trove, I took the opportunity to tweet back quotes from the book. I was bantering with colleagues and thus using humour to create an engaging ‘voice’ for Trove’s Twitter account. I was having a wonderful time. I paused to ask myself, am I just talking to people like me?

Participating in social media spaces is now business as usual for staff at the National Library of Australia, as it is for many organisations in the Galleries, Libraries, Archives and Museums (GLAM) sector. Library staff do this to promote our collections, build brand recognition, build communities and encourage use. Library staff need to analyse how much of each of these is achieved, and with which social media tools. As Delitt and Schindeler noted in 2012, social media tools might be ‘free’ in the sense that they are built and maintained by others, but their use requires resources. (2012, p.12) Those resources should only be allocated if they will contribute to the achievement of the organisation’s mission.

Reporting, like the poor, will always be with us. This is a good thing in a publicly funded organisation like the Library. Feeling strongly that the Library contributes to the public good and makes a valuable impact on people’s lives is not enough. It must be demonstrated to the people who provide the Library’s resources. The Library has for many years used a balanced scorecard approach to report against its strategic directions. For social media activities, this has meant reporting a series of numbers. What do these numbers mean? Do they tell a story of lives enriched and changed?

The framework: strategies and policies that govern Trove’s use of social media

The National Library of Australia, the home of Trove, uses social media tools. The Library’s Corporate Plan 2015-2019 (Corporate Plan 2015-2019) sets out the Library’s agenda for the coming years. Strategic Direction two is ‘Make access happen’. In today’s connected world, users expect that access will happen online as well as within the building. In addition to its own online services, the Library uses third party social media tools, including Facebook, Twitter, Instagram, blogs and email lists. In addition, Trove contributes to this broader social media effort with its own dedicated instance of Twitter. The Library’s Corporate Plan sets out specific goals for the next four years, with several of these directly speaking to staff use of social media:

  • Transform Australian lives by connecting communities to their national collection through vibrant onsite and online programs and services, developed with strategic partners and harnessing new opportunities in the digital environment.
  • Create value for library services and their patrons—across all jurisdictions, regions and sectors—by managing the research infrastructure that underpins access to the nation’s library collections.
  • Extend this value to the gallery, archives, museum and research sectors through sustainable collaboration enhancing access to, use of and engagement with Australia’s cultural and research collections. 

This last point is particularly pertinent for the Library’s use of social media. Connecting and partnering are collaboration: working together towards a shared purpose. Implicit in this is the idea that there will be an exchange of ideas, information or skills that will achieve a beneficial impact. Library staff like to describe Trove as a collection of collections. The Library’s purpose is to share those collections with all Australians. In Trove there is the Library’s collection, alongside a wide range of collections, from small historical societies like the Berrima Historical Society through to large collections like the National Museum of Australia. Trove is collaboration. It is both expected and natural that Library staff would collaborate in social media.

Social media has changed the way communities contribute to and share information with us as organisations and with each other. (Mansfield et al. 2014, p.16) To achieve this, social media has to be about more than simply pushing information at people. In Trove, staff work to make social media just such an exchange.

Trove’s social media communications are guided by the Library’s Social Media Policy (Web Publishing Branch 2015), which includes a set of guidelines that govern the use of Facebook, Twitter, the Library blog and the email lists. The guidelines say that the Library should achieve the following:

  • Generate public awareness of the Library’s services, collections and activities
  • Manage the Library’s reputation and strengthen its brand
  • Build relationships through engagement with community

Library staff believe this is what is achieved with Twitter. Do the numbers support this?

The tools and their use: Trove’s social media voice

The Library aims to increase awareness and usage of Trove; Library staff aim to grow the communities that have gathered around Trove; and Library staff aim to find communities that have no awareness of Trove and what it could offer them. Who are these communities? Trove has undertaken two user evaluation surveys, one in 2013 and a follow up in 2015 in part to answer that question. Two-thirds of respondents use Trove for family history research (Ayres, 2014). A significant proportion of respondents was in the library and information profession, and used Trove to assist others with their research. Sixteen percent of respondents used Trove specifically to correct newspaper text. The 2013 survey found that Trove’s typical user is female, middle-class, 50 years of age and interested in family history. (Ayres, 2014)

The communities Trove is not engaging, or not engaging proportionally to their presence in the population, include teachers, students, people of non-English speaking background (NESB) and Aboriginal and Torres Strait Islanders. (Ayres, 2014) Library staff had assumed that Trove, with its emphasis on unique Australian content, would be a natural fit for teachers and students, and yet that assumption has not translated into usage. The Library’s strategic plan explicitly states that the Library should make access happen for ‘all Australians’, and NESB communities are Australians the Library wants to be reaching but are not. A second user evaluation was commissioned in 2015 to explore why teachers, students and NESB communities were not using Trove. One of the answers to this turned out to be that these groups do not know about Trove. Twitter gives the impression that the Library is engaging with many people, but the echo-chamber effect of the platform is deceptive. The survey revealed that teachers and students are much more likely to be active users of Facebook than Twitter. However, Trove currently uses Twitter and not Facebook.

The Library also seeks to foster collaboration with strategic partners in the gallery, archives, museum and research sectors within Australia and internationally. Trove itself is a collaborative exercise; hundreds of collections from around Australia brought together in one platform. In order to grow Trove, it is not enough to simply ‘build it and they will come’; Library staff need to actively seek new partnerships. It is also important that the Library’s content partners gain value by collaborating with the Library through Trove; promoting content partner collections is one of the ways Library staff hope to add value for them.

The use of Twitter as an outreach tool for Trove began in 2011, and the service now has an established voice. By mid-2015 Trove’s twitter profile had over 10,000 followers.  Studies of successful organisational Twitter profiles have found that a successful voice is engaged, excited and above all, ‘human’. (King 2015, p.16-17; Shively 2015) Staff in Trove have adopted a consciously more informal voice than the Library’s @nlagovau Twitter profile. Trove’s Twitter voice is also produced by several members of the team, who are encouraged, under general guidance, to tweet about topics of interest to them and in ways that are comfortable and natural to them. Enquiries to Trove will not infrequently arrive with introductions like ‘I know Trove is just a big machine…’ With the voice that has been adopted for Twitter, Library staff endeavour to counteract that impression. The Trove team is small: nine people tweet for Trove, and it is just one of a variety of tasks for which the team is responsible.

Creating content that works

In 2014’s collection of articles reflecting on Twitter, Twitter and Society, the different types of tweet are discussed in the context of Twitter’s evolution. (Weller et al. 2014, p.xii-xxiii) Tweets evolved from banal daily chatter, through conversation, to sharing information and reporting news. More recently, Twitter has evolved into a forum for grass-roots activism. (Weller et al. 2014, p.xvii) For GLAM organisations, an obvious use for tweets is to promote ‘news’, including new acquisitions, events, opening hours, service interruptions and search tips. It is all readily digestible information an organisation can ‘push’ out there. In the interests of ‘writing once, publishing multiple times’ the Library will use each social media tool to cross-promote content: for instance, blog posts will be promoted in Twitter.

Analyses of the topics regularly tweeted by the Trove team suggest that the frenetic nature of Twitter is best captured with the iconic, the unusual or the cute, or issues which readily fit into the current debate. Historian Rebecca Onion said of the history which catches attention on the web that, ‘I think the kind of history that does well on the web is the stuff that’s really unambiguous.’ (Loeb 2015) These tweets will encourage click throughs and ‘engagement’. Information sharing, collection promotion and collaboration can be a slower burn. Here are examples of Trove’s go-to tweet topics: new collection promotion, information sharing and collection fishing. 

What’s new in Trove: collection promotion 

Business as usual for the Trove team includes adding the metadata for collections from Australian GLAM and research organisations to Trove. It is our standard practice that each new collection will be promoted in Twitter once the records have gone live. Examples include the promotion of a new Victorian Collections partner, the Australian Wool Museum. The tweet promoting this collection received a total of 2,655 impressions, including 14 retweets and 11 clicks on the link.

It’s all about food: collection fishing

A popular method for seeking click-throughs into Trove is to ‘fish’ for random collection items and tweet links to those items. This is called ‘collection fishing’. One of the most popular tweets of the first six months of 2015 was about Vegemite. Trove uses hashtags sparingly (Shively 2015), and endeavours to capitalise on existing ones to tap into established conversations. The #onthisday hashtag is one of those, and on 13 June at 6.30 on a Saturday evening, a tweet celebrated the anniversary of Vegemite’s launch in 1923 by including an advertisement from Trove’s digitised newspapers. Within 48 hours, this tweet had been re-tweeted 82 times and ‘favourited’ 32 times. There were 12,155 impressions, 231 clicks on the attached image, 90 clicks through to Trove, 26 profile clicks and we gained two more followers.

Open access: information sharing

One of the most important agendas the Library has been pursuing over the last year is to expose open access content to users through Trove. Time and again, user expectation surveys indicate that people want digital content now, and Library staff have been keen to help them find it. On 25 May this year, Library staff tweeted tips for searching for images that had Creative Commons licenses, which meant that the images could therefore be safely reused. This tweet was ranked 20th by Great Oz Gov tweets for that day. It was retweeted 33 times and favourited 21 times. There were 5,878 impressions and a total of 151 engagements, which includes link and profile clicks. This tweet did result in a reasonable number of link clicks, at 56.

How do I know who I am speaking to and if they are listening?

Collecting numbers for the sake of it is a meaningless activity. As David Lee King says, ‘tracking and recording monthly numbers is easy. Actually wringing meaning out of those numbers can be where the challenge lies.’ (King 2015, p.29) Ideally, metrics provide information to demonstrate value and inform decision-making. (Konkiel, Dalmau & Scherer 2015) The author's thinking about the value of the social media metrics collected by the Trove team is informed by the impact analysis work of Simon Tanner, especially his definition of impact: ‘The measurable outcomes arising from the existence of a digital resource that demonstrate a change in the life or life opportunities of the community for which the resource is intended.’(Tanner 2012, p.4) While this model relates to the use of a digital resource, there is value in considering impact in the context of social media metrics. Impact requires that a change or action has resulted thanks to the encounter with a resource, or in this case, tweet. Can the numbers truly measure impact?

Library staff use Google analytics and Twitter’s analytics to gather performance metrics. These services are free and easily accessible. Reporting in the public sector has often defaulted to ‘quantitative performance indicators…where very basic metrics…remain pre-eminent as proxies for qualitative experiences.’ (Tanner 2012, p.26) Library staff are collecting both what is easy and what is expected, and go on to make assumptions that this translates into impact.

Library staff collect the following quantitative metrics for all social media accounts, including Trove’s:

  • Total followers in Twitter
  • Total ‘likes’ of the Facebook account
  • Total page views of Flickr Commons
  • Total followers on the Instagram account

Other measures for each platform include:

  • Reach: total followers
  • Activity: new tweets
  • Engagement: new direct mentions and replies
  • Amplification: retweets
  • Conversion: click throughs

Each platform reflects these measures in different ways. Metrics mappings for social media outline where each measure is taken from in each platform.

  Twitter Facebook Blogs Flickr Mass email
Reach Followers Likes Page views Views Subscribers
Activity Tweets Posts Comments Uploads Posts
Engagement Mentions Comments Comments Comments Opens
Amplification Retweets Shares      
Conversion Conversions Conversions Conversions Conversions Conversions

Table 1: Metrics mappings for social media

Trove’s management reports twice yearly to the Trove Reference Group on the following measures for Trove’s Twitter profile, and Trove’s blog posts, which are a stream in the Library’s blog platform:

  • Number of tweets published (Activity)
  • Tweet impressions
  • Tweet engagements
  • Click-throughs
  • Number of blog posts published
  • Number of blog post page views
  • Top tweets compared to Australian government agencies

"The Trove Reference Group provides advice on the medium term strategies and priorities for Trove’s development. Members are drawn from each division within the Library."

Twitter analytics provides the following measures:

  • Impressions: the number of times users saw the tweet on Twitter
  • Engagements: total number of times a user has interacted with a tweet, including clicks on hashtags, avatars, username, tweet expansion
  • Engagement rate: number of engagements divided by the number of impressions

Engagement sounds like a very promising measure of impact. Engagement involves the tweet viewer actually doing something: clicking on some part of the tweet. An action has been inspired by the existence of the tweet. This number does not indicate, though, whether their lives were enriched by what they saw or if they just continued to swipe through their feed. The number doesn’t reveal what happened next. 

Impressions can be a very impressive number, particularly when high-profile people retweet something from your service. Suddenly there are many people ‘seeing’ the tweet. Impressions cannot reveal, however, if they have read it, or whether they have only scrolled through their feed without noticing the content. Equally, the number of followers of an account, a number any organisation enjoys seeing rise and reports faithfully, only really says that someone was interested enough to click the follow button. As Derek Thompson puts it, ‘People read without sharing, but just as often, perhaps, they share without reading.’ (Thompson 2015) 

A better story of impact is told by the qualitative data: stories from Trove’s users themselves, gathered from conversations in Twitter, comments on the Trove blog, from the Library’s online enquiry system and the commissioned user surveys. This feedback is users telling the Library that Trove has made an impact on their lives and what that impact is. These qualitative data are reported alongside the quantitative measures. Feedback is, by its very nature, much more resource consuming to collect and report on meaningfully, as it requires staff time to gather, rather than the numbers, which are gathered by an algorithm. Moreover, apart from the commissioned surveys, the majority of this feedback is serendipitous. It adds colour and depth to the numbers.

What it all means: collaboration both in and beyond Twitter

In late 2014, three Chinese language newspapers were added to the digitised newspapers zone. The Library considered these to be an important step forward in representing Australia’s Chinese heritage, and the inclusion of these newspapers was promoted heavily. Trove staff invited the historian Kate Bagnall, who specialises in Chinese-Australian history, to write a guest blog post about these newspapers. (Bagnall 2015) The post was promoted through Twitter by Trove, the author, and the Library’s main Twitter account. The post and the Chinese newspapers were also promoted through the Asian Australian Studies Research Network and the East Asian Library Resources Group of Australia, as well as being reported in the PS News Online (Chinese whispers given new voice 2015) and in an article written for Incite (2015) by Tim Sherratt. This post was the most viewed blog post for the period January to June 2015, with 860 views. The use of a guest author for a Trove blog post, and thus a different voice speaking for Trove, allowed the Library to tap into a community of historians that may otherwise never have been reached.

The numbers for the following outreach effort tell the sort of impact story in which Library staff are most interested. One of the priorities for Trove for the last eighteen months has been to increase the availability of open access research. Achieving this has involved engaging actively with the research sector. Trove staff wrote two blog posts as part of these efforts, and they were promoted in Twitter in both the Trove and the Library accounts. ‘Spot the blue square: helping you find Australian researchers’ (Hickie 2015) and ‘Untangling access: Helping you find freely available content online’ (Sherratt 2014) were both posts which appeared on the Trove blog between January and June 2015. They received 553 and 323 page views respectively. This places them comparatively high within the ten blog posts published in this period, but neither was the most successful, and the numbers of views are dwarfed by the views of the falling dog article. The outcomes of these posts, however, met the Library’s aims very effectively. Both posts were shared with the research community using Twitter and industry-specific mailing lists. As a direct result, Trove staff were contacted by members of the research sector with new content for Trove, improved metadata for records already in Trove and requests for improved data in Trove. This did indeed allow Trove to lead, partner and connect with an important strategic partner. To paraphrase Tanner, a measurable outcome was achieved by this outreach. (2012, p.4)

On two occasions during August 2015, Twitter engagement provided evidence of impact: the Trove Twitter community introduced Trove to members of their own communities and lives were changed. On 6 August, a PhD student with the Twitter handle ‘PhD hater’ posed the following question to her Twitter followers:

‘Sydney! Does anyone have a copy of ‘Violence’ by Wolfgang Sofsky that I can borrow for a week or two? You’d be helping me heaps’ "this tweet has been deleted from twitter". Trove followers Stephen Murray and Liam Hogan tracked down the work for her, telling her what Trove is. Liam Hogan said, ‘mate book yourself a day or two and get around trove.nla.gov.au @TroveAustralia’. PhD Hater’s response, while profane, indicates impact, ‘fuck this is brilliant’ "this tweet has been deleted from twitter". Another post-graduate student discovered Trove as part of a subject she was studying and her response screams impact, ‘Just discovered @TroveAustralia my life has changed forever. Thanks #isys1166.’

Trove’s Twitter community

I have said that I want to know if I am just talking to people like me. So who am I?

The Australian Library and Information Association Education, Skills and Employment Trend Report 2015 (Skills and employment trend report 2015) notes that employees in this sector are between six and ten years older than the median age for all occupations, with a median age of 49 for librarians and 46 for library technicians. (Skills and employment trend report 2015, p.14) Library and information professionals had attained higher education qualifications at a higher rate when compared with all other professionals: 60% had a Post Graduate or Graduate Diploma or Graduate certificate. (Skills and employment trend report 2015, p.18)  People working in the library and information world are older and educated.

In this context, I am also part of the GLAM sector. The GLAM sector ‘ranges from the very large to the very small, from capital city to remote settlement, from well-resourced to no resources, from paid to volunteer’. (Mansfield 2014, p.2) These sectors are often referred to as the collecting organisations: they are, primarily, non-profit, receiving all or most of their resources from government. (Mansfield 2014, p.2)

To gain insight into Trove’s Twitter followers, we took advantage of Twitter’s application programming interface (API) and the analytics service this provides. The analytics service had its limitations, in that it only offered a very limited range of filters by which to slice and dice user data: the telecommunications companies used by followers, the television genres they are interested in, and their general interests. It was not able to provide information about the number of corporate versus individual accounts, occupation or where followers live. For this data, we turned to information gathered from the API.

Over the first six months of 2015, Trove gained 1191 followers. When analysis of Trove’s Twitter followers began, we had 10,206 followers. Initial attempts to gather follower data involved combing through the follower list on the web. When this proved to be both time consuming and ineffective, the Twitter API was used to harvest the user information. This information was downloaded into a spreadsheet. From the initial spreadsheet, every 99th user was identified and removed to the top of the spreadsheet.  In order to have a controlled test we then took every 99th username from the same list, minus the ones initially extracted. This process jumbled the names to ensure randomisation. The two samples were added together, providing two percent of Trove followers for analysis. The key fields provided by the API used for the analysis included ‘stated location’, ‘total tweets’, ‘bio’, ‘major interests’ and ‘about’. Not all users selected for study have entered information in each of these areas. Where they have done so they are using free text rather than a controlled vocabulary, and some interpretation of the information was necessary. On occasion, choice of username did not make it clear whether the account was that of an individual, organization or Twitter bot. In these cases a best guess was made. 

This analysis was pursued to uncover:

  • Where do users come from?
  • What is the typical occupation of a follower?
  • How many corporate or group accounts follow Trove? 

In essence, are our followers just like me, a middle-aged librarian with a certain sense of humour, employed by a leading cultural organisation? If they are just like me, how far will further outreach in this medium help Trove to reach the sectors where Trove awareness and usage is low?

Of the 202 Twitter followers studied, 164 were individuals and 38 were organisations.

 Proportion of Trove's followers that are organisation 29% yes 71% no

The majority of followers who included a location are Australian: 97 list Australia or an Australian location, 17 another country and 51 had no location identified.

 Nationalities graph. Australia 97, United Kingdom 5, United States 3, Indonesia 2, Phillipines 1, No stated 51, Honk Kong 1, Malaysia 2, Tunisia 1, New Zealand 1

Information about professions and interests was spread over the ‘bio’, ‘major interests’ and ‘about’ sections. Of the 43 sampled which list a profession, 15 clearly identified themselves as GLAM sector employees. The next highest type of profession was education, which included one teacher and three academic disciplines. Other professions represented, included media, management, marketing, health and information technology.

 Professions graph. GLAM 15, IT 4, Arts 5, Education 4, Health 2, Management 3, Media 2, Marketing 2, Law 1, Finance 1, Science 1, other 1

Information about follower interests and hobbies was comparatively sparse. There were twenty-three followers who included information in these fields. Twelve had interests in various forms of the arts, with politics and social justice, sport and religion popular topics of interest. 

 Interests graph. Religion 4, Arts 12, Sport 3, Politics/social justice 5, family history 2

Twitter’s analytics service provides some information about a profile’s followers which tallies with this analysis. Based on the information in the table below, Trove’s followers are, in comparison with all Twitter followers, more interested in culture, current affairs and science.

Interest name Trove’s followers Twitter followers
Books news and general information 84% 12%
Business and news 77% 38%
Politics and current events 73% 33%
Movie news and general info 70% 47%
Business news and general info 64% 26%
Science news 62% 13%
Comedy (movies and television) 58% 69%
Music news and general info 54% 41%
Technology 53% 21%
Tech news 43% 19%

Table 2: Twitter analytics captured 9 September 2015

The majority of the organisations following Trove in this sample are GLAM sector organisations, with 16 identifiably from this sector. There were six genealogy groups, one education group and 14 from organisations as varied as a timber flooring company, bookshops, publishers and a Cairns dive school. 

 Organisations, GLAM 43%, Genealogy 16%, Other 38%, Education 3%

We now know that most of Trove’s twitter followers are Australian, their typical occupation and interests are in the GLAM sector and the arts, and nearly a third of followers are organisations. It is therefore very likely that I am, for the most part, talking to people like me. This meets the Library’s aim of partnering and collaborating with peers, but it is not reaching out to new audiences or building new communities. The Library needs to be looking elsewhere for them.

Conclusion

If driving traffic to the Trove website were all Library staff were trying to achieve with a social media presence, then it would be far more efficient to depart the field and find a way of regularly seeding Reddit with weird and wonderful articles. As Thompson says, ‘The more sophisticated takeaway is that Twitter is worthless for the limited purpose of driving traffic to your website, because Twitter is not a portal for outbound links, but rather a homepage for self-contained pictures and observations.’ (Thompson 2015) What Library staff outreach does achieve is a great deal of partnering and connecting. The Library is in the space fellow GLAM organisations are in, and as such staff are in a good position to work with colleagues there. However, there is more to do to reach the people, communities and industries who are not like us, and Twitter probably cannot take us there. The Library will continue to count who, what, when and why. Library staff will continue to explore ways to collect and measure the impact of Trove through these social media. Numbers, however, are too blunt an instrument to demonstrate what we as an organisation really want to know.

In June 2014, Trove received a compliment through the Library’s online enquiry system. A man, who grew up in a children’s home, commented to his daughter that he remembered a journalist taking his photo when he was ten years old. Within a few minutes, his daughter had searched Trove and found that photo, the only one taken of this man during his childhood. Trove returned a piece of someone’s past, a gift that a number cannot measure. Library staff want to know how wide an audience is reached using the numbers. Library staff want to engage with communities and measure that through engagements and impressions. Above all, Library staff want to be able to collect and share the stories of impact, where lives have been enriched and changed by the work the Library does. In answer to the question this paper poses, does a click-through mean what we think it means, the answer is that it does not answer the questions we want to ask: what impact do Library services have on the community. In late 2015, the Library will be exploring ways to capture systematically these impact stories in order to report against key performance indicators. Library staff will know, and be able to demonstrate, that we are fulfilling our role as the nation’s memory, fostering ideas, creativity and a passion for learning.

Acknowledgements

Thanks to Chris McPartland and Tighearnan Kelly for their contributions to the statistics gathering and analysis.

References

Australian LIS Education, Skills and Employment Trend Report 2015Canberra ACT, Australian Library and Information Association, August 2015 [8 December 2015]

Ayres, Marie-Louise, 2014, Trove at 5: are we there yet, NLA staff papers [8 December 2015]

‘Chinese whispers given new voice’, PSnews online, no. 44, 24 February 2015. [8 December 2015]

Corporate plan 2015-2019, National Library of Australia, [8 December 2015]

Dellit, Alison and Schindeler, Sarah, 2012, ‘Trove: the terrors and triumphs of service-based social media’, VALA 2012 eM-powering eFutures. [8 December 2015]

King, David Lee, 2015, Managing your library’s social media channels, Library Technology Reports, vol.51, no.1. January 2015

Konkiel, Stacy, Dalmau, Michelle, Scherer, David (2015), Altmetrics and analytics for digital special collections and institutional repositories, Digital Collections Services, Indiana University [8 December 2015]

Loeb, Eryn, Beyond the Archives: An Interview with Rebecca Onion, 2015, [8 December 2015]

Mansfield, T., Winter, C., Griffith, C., Dockerty, A., Brown, T. 2014 Innovation Study: Challenges and Opportunities for Australia’s Galleries, Libraries, Archives and Museums, Australian Centre for Broadband Innovation, CSIRO and Smart Services Cooperative Research Centre, August 2014, [8 December 2015]

Sherratt, Tim, 2014, ‘Growing together: Trove and Victorian Collections’, [8 December 2015]

Sherratt, Tim, 2015, ‘New perspectives on Trove newspapers’, Incite, vol. 36, no. 6/7, Jun/Jul 2015, 19.

Shively, Kevin, 2015, ‘5 surprising tips for tweets that drive engagement’, Simply Measured, [8 December 2015]

Tanner, S. 2012, Measuring the Impact of Digital Resources: The Balanced Value Impact Model, King’s College London, October 2012

Thompson, Derek, 2015, ‘The unbearable lightness of tweeting’, The Atlantic, [8 December 2015]

Web Publishing Branch 2015, Social Media Policy, National Library of Australia, [8 December 2015]

Weller, Katrin, Bruns, Axel, Burgess, Jean E., Mahrt, Merja, Puschmann, Cornelius 2014, Twitter and society : an introduction. Peter Lang Publishing 

Appendices

Appendix A

Google analytics for the article Three dead after dog falls 13 floors (1988, October 24). The Canberra Times (ACT : 1926 - 1995)

Page Pageview Unique page views Avg. time Entrances Bounce rate Exit
/ndp/del/article/102017012 253621 244388 0.001261574 244319 0.9648 0.9616

Day index

Date Page views
5/06/14 0
5/07/14 0
5/13/14 180,774
5/14/14 51,701
5/15/14 2,169
5/16/14 952
5/17/14 727
5/18/14 802
5/19/14 737
5/20/14 404
5/21/14 259
5/22/14 220
5/23/14 201
5/24/14 157
5/25/14 199
5/26/14 245
5/27/14 174
5/28/14 218
5/29/14 165
5/30/14 296
5/31/14 171

Page view numbers after this date rapidly become double and single figures and trail off almost completely by the end of July 2015. 

Appendix B

Most popular tweets as ranked by Great Oz Gov tweets, January to June 2015

Date Subject Ranking Retweets Favourites Reach
15/02/16 NSW flag 2 34 15 83021
14/04/16 Cat kills stingray 6 21 9 67928
9/02/16 Kingsford Smith 20 12 6 67221
17/02/16 Fundraising campaign 7 21 8 65185
28/05/16 Title no. 930: The All British 48 12 3 58933
17/03/16 St Patrick’s Day 15 15 10 52726
18/05/16 International Museum Day 34 12 5 52081
3/06/16 Mabo decision 35 15 2 51267
8/05/16 David Attenborough 3 15 16 50700
25/06/16 Asleep in the Sydney Domain 7 15 13 48853
13/02/16 Top 25 newspaper articles of 2014 42 14 4 48610
1/06/16 Sleds on Mt Buffalo 47 9 4 47879
13/06/16 Vegemite 1 32 16 45948
23/06/16 Cow with wooden leg 41 9 5 45921
2/06/16 Italian National Day 27 14 3 45816
1/04/16 April Fools’ Day 39 13 3 45606
3/02/16 Storify of Trove at ALIA Online 11 14 12 45354
27/04/16 Laser guided robots 24 11 7 44315
2/03/16 Dr Seuss 8 20 13 43436
19/03/16 Sydney Harbour Bridge opening 9 21 16 42717
22/06/16 10,000 followers 22 5 10 40443
13/01/16 New newspapers 36 9 6 39900
26/03/16 Steam engine for Harbour Bridge 43 8 8 39120
27/02/16 Polar bear escape 40 8 6 37792
27/03/16 Visiting Trove in your pjs 37 5 8 37577
6/01/16 Trove bugs 48 6 5 34597
5/01/16 Shearers’ strike 1891 41 11 5 33386
9/06/16 Crossing of the Pacific by Southern Cross 21 13 5 28084
8/04/16 Big Things 16 17 6 26783
5/02/16 Welcome Stranger 22 12 9 26461
27/05/16 1967 Referendum 49 12 4 25234
10/02/16 Teams behind Trove 17 12 11 24044
2/02/16 Superb owl 4 22 18 23024
25/05/16 Creative commons licensed images  20 14 11 21836
27/06/16 Queenslander covers 12 8 6 19503
10/03/16 Cake for Twitter followers 9 15 11 18999
1/05/16 National Wool Museum 28 10 8 18643
2/04/16 Tram conductors 45 6 6 18486
21/02/16 State flags 43 11 7 16918
21/06/16 Text correction tips 29 5 8 14266
5/05/16 Kitten and camera 36 6 8 13987
13/04/16 Title no. 874: Melbourne Punch 38 8 6 13746
6/05/16 Meteorological observatory on Kosciuszko 38 9 5 13434
15/01/16 Ladies tennis 47 7 4 13304
20/06/16 $2 coin 36 6 7 13255
19/06/16 Latest newspapers 50 5 6 12746
8/03/16 International Women’s Day 41 10 4 12106
27/01/16 Jackie French 29 8 13 11697
10/06/16 Houdini in Melbourne 44 6 5 11544
30/01/16 Colleen McCullough 21 7 8 11213
16/03/16 9000 followers 13 5 15 10841
12/02/16 Book Lovers’ Library 48 9 5 10545
17/04/16 Trove article in bus shelter 45 2 9 10285
10/05/16 Mothers’ Day 42 3 7 10263
14/02/16 A valentine 50 4 4 9652