Alison Dellit
Sarah Schindeler
Publication date: 
Sunday, 1 January, 2012

The National Library of Australia has actively used a range of social media tools for the promotion, development and delivery of services for a number of years. In addition to whole-of-library branded activity that is managed centrally, teams elsewhere in the Library are creating niche and service-based social media channels. Using the 2011 trial of the Trove social media rollout as a case study, this paper examines the rewards and challenges associated with niche or specialised social media engagement, as well as the broader, potential implications for online engagement by cultural institutions.

The Context

Going out to meet the people

In recent years, the coverage and influence of social media services has grown exponentially. For the purposes of this paper, social media can be broadly defined as digital services that enable users to perform tasks, interact, and be entertained through establishing relationships with other users of the service. This definition is an adaptation from Kaplan and Haenlein (2009). In addition to the burgeoning number of dedicated social networking services, social media features have also been integrated into services that are otherwise simply online channels for broadcasting content. For example, Flickr allows users to upload pictures online and also allows users to comment and tag others pictures, join groups and interact and share content with other users. Other services with social networking components include, amongst others, Facebook, Twitter, MySpace, LinkedIn, Tumblr and TripAdvisor.

In the last 12 months alone, Neilsen research (2011) found that the proportion of online time users spend on social media has increased by 43% and the trend shows no signs of slowing. The same study also found that amongst the key online markets globally, Australian Internet users spend the most time visiting social networks and blogs, averaging 7 hours, 17 minutes per person a week (Neilsen 2011).

Further, not only are users spending more time online, there is a steady stream of new users to social media services. In the month of October 2011 alone, the Australian Facebook population grew by over 30,000 (new people registering accounts), LinkedIn experienced a significant jump internationally, and Twitter reported a significant increase in Australian visitors to its site (SocialMediaNews, 2011). Any earlier scepticism regarding the relevance and endurance of social media is waning. Both commercial and public entities are recognising that social media is now replacing email as a dominant tool for personal communication and represents a highly effective means to reach our communities.

Where cultural and research institutions stand

In the early days of social networking, it was considered “personal” activity and unrelated to professional life or the corporate world. In the last five years, however, engaging in social media has become an essential tool for organisations and companies.

There are several factors contributing to this:

  • Social networking sites provide high-value word-of-mouth marketing, as people trust recommendations from individuals they have selected to be part of their circle;
  • Social networking services allow organisations to go into spaces people, particularly young people, already inhabit, instead of relying on them to come to a corporate space. It is the online equivalent of handing out condoms in the nightclub, rather than offering them at a health centre across town; and
  • Without engagement with social media, organisations are not able to influence conversations about themselves. From a reputation management perspective, this includes rectifying problems and correcting erroneous information.
  • Many organisations join in recognition of the growing influence of social media and hope to learn by doing, at times without identifying specific goals or objectives.

Unsurprisingly, cultural and research institutions have engaged in differing ways, and to differing degrees with social media. These institutions use social media to perform different tasks, including the examples listed below.

Informing the public about exhibitions, events and offers

This is the most common use of social media by cultural organisations, particularly among arts organisations. Institutions with high profile exhibitions, such as art galleries tend to have high numbers of followers for such activities.

Drawing attention to collection material

Good examples of this include the Twitter feeds of the Smithsonian and the New York Public Library, both of which feature items relevant to current events and historical anniversaries.

Creating communities of users

A few cultural organisations have integrated social networking features into their programs in order to facilitate interaction between users. A great example is Washington State University's Plateau People's Web Portal, which enables users to register profiles, connecting with other users and establishing tribal identities, which can provide access to restricted resources. A more common example is OCLC WorldCat functions to allow private profiles. The leading cultural activity in this area, however, has been set by commercial websites, such as LibraryThing and Goodreads, which allow users to create virtual collections and develop communities around them. Within Australia there are a number of other institutions also have online communities, including Trove with its forum and the online community of the Australian Museum.

Providing leadership on issues pertinent to the sector

In some sectors, opinion leaders within organisations maintain Twitter feeds to comment on sectoral issues. Unlike many feeds of influential individuals, these feeds are focused on professional issues and clearly identified with the individual's employer. Examples of such feeds are those run by senior OCLC staff, such as Roy Tennant and Lorcan Dempsey. A good Australian example from the media sector is the Twitter feed of ABC Managing Director Mark Scott. Feeds such as these have become increasingly influential.

Crowdsourcing of information

There is a high level of activity in this area in the last 10 years in the cultural sector, from projects such as Pennsylvania State University's Penn Tags through to Kew Garden's Flickr project. Trove has established itself as a world leader in this area, through the newspaper correction program, and it also contains an image contribution feature utilising Flickr.
Social networking services can facilitate participation and contribution to the work of an organisation. The aforementioned “crowdsourcing” of content allows institutions to enrich their collections in ways previously not possible.

Online profile via search engines

In 2011, emerging analysis shows that most major Internet search engines now factor social media discussion topics into their search result ranking algorithms, ensuring that organisations with a strong social media presence will also have a strong Internet profile (Fishkin 2011).

Snapshot of social media at the National Library

The National Library of Australia has experience using social media tools for the promotion, development and delivery of services. The earliest activities aimed to engage with communities of users to crowdsource content and commentary on our collections. In 2006, a collaborative wiki space called Take Part was launched as part of the Australia Dancing service, to enable dance enthusiasts to contribute entries about dancers, companies and organisations. Shortly afterwards, a folklore wiki was created to encourage the posting of items and comments about the Library’s folk collections. Despite strong collections and strong communities of interest in these two areas, neither initiative attracted any significant user participation.

However, one of the success stories from this same period was the Library’s use of Flickr, an online image repository for photo storage and sharing, as a means to increase the number of contemporary images in PictureAustralia. Previously PictureAustralia gave users access to images only from collecting institutions such as libraries, galleries archives and museums. Flickr was used as an instrument to receive photographic submissions from the public. The PictureAustralia Flickr group (now feeding into Trove) has 2,836 public members and has accepted nearly 100,000 images of Australia and Australians.

Today the most high-profile Library-branded activities are Twitter and Facebook, with a combined following of some 6,500 accounts. These are used daily to support a range of communications and marketing activities, including:

  • broadcasting news;
  • promoting events, services and collection material; and
  • monitoring and responding to feedback and enquiries.

The Library also has accounts with a number of other social media services, such as YouTube, Flickr and SlideShare. While technically classified as social media tools, based entirely on the platforms used, in most instances these have been used by the Library as ad hoc broadcast channels for instructional and promotional content sharing, such as interviews with Library-published authors, presentation slides and ‘how to use the Library’s services’ screencasts.

In addition to the Library-branded activity managed centrally by the National Library's Communications and Marketing team, staff elsewhere in the Library are creating niche and service-based social media channels. The most recent example is the fringe publishing blog launched by the Collections Management Division in June 2011, which aims to build relationships with the zines community and expose fringe collection material to a wider audience. The blog has quickly gained a strong following reflected in some 5,300 visits. More importantly, it represents a more than successful example of community building and positive engagement - individual blog posts routinely receive online comments from members of the zines community.

The most high profile niche activity was undertaken for the Trove service. In November 2011, the Library began a trial social networking based marketing push, with a focus on Twitter, Facebook and a blog-style posting in the Trove forum.

Concurrent with the Trove team managing a social media trial, the Library began a corporate process to develop a social media strategy. This process identified the following overarching organisational drivers for all social media activity at the Library:

  • Delivering service;
  • Improving online access;
  • Participating in online communities to enhance visibility and reach new audiences;
  • Monitoring and adopting international best practice; and
  • Community engagement and strengthening relationships.

The Library recognised the scope within different operational areas to utilise social media for vastly different purposes, and that initiatives would be, by necessity, quite wide ranging, while still supporting the overarching strategic directions. Different activities would need different goals. In this context, the goals for whole-of-library activity were identified as follows:

  • Generate awareness of services, collections and activities;
  • Reputation Management and brand strengthening;
  • Build relationships through engagement;
  • Research and user insight; and
  • Customer/client support.

Different goals are identified for niche social media activity as proposals arise, although there is likely to be strong alignment between the whole-of-library and niche goals. As a matter of course, any new social media engagement activity at the Library has defined and measurable goals and objectives and an identified audience.

The Case Study

Trove is a user-driven service

Trove is the National Library’s premier search service, providing access to 250,000,000 resources, held by thousands of libraries and other institutions. Trove is a collaborative effort, run and managed by the National Library, but supported by services such as Libraries Australia, networks such as National and State Libraries Australasia, and by thousands of contributing organisations.

This collaboration not only shapes the development of the service but is also embedded in how it operates. Trove is supported by a strong user community, who not only visit, but also participate in the service, adding comments and tags, and improving the searchable newspaper text. Trove is designed to be an interactive service that facilitates both the discovery and the capture and creation of new knowledge.

In particular, the activity of volunteer text correctors has broken new ground for Australian services. Many thousands of people have corrected more than 50 million lines of text, with around 60,000 lines corrected every day. For more details on this work, see Holley (2009). Text correction is the most significant engagement, but this is far from the only way users engage with the system. Every month, thousands of images are uploaded to the system by members of the public, via a Flickr group established for the purpose, and 10,000 lists have been created, providing context to existing material. Trove also maintains an online user forum, which supports conversation between Trove users, and between Trove users and Library staff.

This level of engagement has brought with it a strong sense of ownership of the service by those who use it. Trove has a passionate support base, the members of which rightly expect to have input into the direction of the service. In this context, social media services offer a way for Trove to converse with our users in a peer-based way.

A separate Trove social media presence

Social networks are widely believed to be one of the greatest boons to niche marketers. A ‘savvy’ operator can establish an online community of users, who are given the opportunity to connect and network with like-minded people, and then reap the rewards of tapping into those connections.

The challenge that cultural institutions face is that their services are necessarily cross-market and diverse, which is reflected in the content they broadcast on their social media channels. Consequently, users who are consuming that content are more likely to be subjected to large quantities of irrelevant information and only a proportionally small number of items of interest. This is one of the reasons why so many channels managed by cultural institutions are attracting an audience that is dominated by other cultural institutions and their staff, although this is rarely the intent.

Channels that focus on specialised subject matter may have a smaller pool to draw upon, but the followers will be interested and therefore more likely to engage and be genuinely invested in the relationship. This has been very well illustrated by blogs by institutions such as the New York Public Library, Library of Congress and British Library, all of which deliver highly specialised blogs on topics as wide ranging as inventions and patents, medieval manuscripts, faith and science fiction.

The decision to instigate a Trove-specific social media presence alongside the National Library’s media presence was shaped by a discussion about the potentially specific aims and objectives that could be achieved by such a feed, and also on whether a specific audience for a Trove feed existed.

The nlagovau Twitter feed, @nlagovau, for example, promotes a lot of events and exhibitions in the building (one of the most successful uses for a cultural organisation Twitter feed). As most of @TroveAustralia’s followers are not from Canberra, and do not identify with a particular institution, “cluttering up” the feed with such information is just as annoying as regular readers being pelted with Trove results every day.

Additionally, a basic principle of social media is that contact should be as unmediated as possible. This is particularly true of Twitter: a network within which an hour can be a lifetime. The Trove twitter feed is run and managed by the staff working on the Trove Support Team. By having the staff most familiar with the service able to answer questions, or disseminate information, immediately and authoritatively, a minimum of double handling was achieved.

The decision to use social media tools to communicate with Trove users was driven in part by its existing passionate and active user-base, one which seemed likely to want to actively engage with Trove staff in order to better inform the service. For the Trove Support Team, using Twitter as a marketing mechanism had several advantages over other forms of marketing activity:

  • Social media provided an opportunity to directly showcase the content of Trove, not just the interface functionality;
  • The necessary skills could be picked up by existing staff members, making it more cost-effective, as well as creating a strong connection between the staff developing the service and the marketing; and
  • Word-of-mouth was already a significant driver of new Trove users, and social media provided an opportunity to capitalise on that.

It is important to note that with limited resources, no broader publicity for any of these features was undertaken, and all were launched very softly.

Diving into the pool

In accordance with the National Library’s overall social media drivers, the Trove social media activity set aims for the work:

  • To increase use of Trove;
  • To increase the visibility of Trove;
  • To provide customer service to Trove; and
  • To solicit feedback about Trove, in order to improve the service.

During the first six months of the activity, two outcomes were added:

  • Request feedback about specific Trove features or items; and
  • Respond to questions and comments about Trove.

With very little practical experience of social media on the Trove Support Team (a relatively small group of five staff), the initial plan was to use three different social-media channels to distribute common content. This would start with the construction of a short forum post by a member of the Trove Support Team, focused on an item of interest that could be highlighted by Trove. This might be inspired by items in the news, popular culture, or by significant (or not so significant) anniversaries.

The desired outcomes were identified as follows:

  • Highlight Trove collection items, with a focus on quirky and undiscovered collection content;
  • Provide timely updates about developments and technical issues/closures; and
  • Demonstrate ways in which the resources of Trove have relevance to topical and everyday news and activities, outside of the traditional formalised research scope.

Once a forum post had been written, the content would be packaged into a short post on the Facebook page, and a short tweet would be created to link to the content.

As the team learned, the social media landscape is vast, and fairly complex. Every social media application is different, some dramatically so, and things did not turn out as expected.

Learning 1: Not all social media is the same

The quickest reality check faced by the team was the very different nature of Twitter, Facebook and blogging. The forum-post content was the most resource-intensive to create, and turned out to be difficult to adapt for use in other channels. For example, the size restrictions placed on Facebook posts necessitated simplification of the posts, limiting their usefulness considerably. In addition, attaching multiple URLs to a Facebook post posed significant challenges. At the same time, short tweets pointing directly to Trove proved more popular, and faster and easier to create, than those pointing to a forum post, which then pointed back to Trove.

Twitter displays some very different characteristics to Facebook. Firstly, the level of reciprocity between accounts is significantly lower: Twitter users (known as tweeps) form much less distinct communities than Facebookers, and hence the degree of separation between any random two users on Twitter is shorter, on average, than the same line on Facebook (Kwak et al 2010). If you think of social media as a game of Whispers (also known as Telephone and Chinese Whispers) using Twitter is like playing in the MCG, whereas Facebook is more like a series of disconnected private parties. It is easier for your message to reach the other side of the MCG than to jump from party to party. An important exception is the tendency of Tweeps to follow and interact with others in their own time zone, a trend less distinct on Facebook. This is expected, given the time-dependent nature of the service.

While Facebookers have an average of 130 friends each, most users interact intensively with a significantly smaller group (Facebook 2011). One study has suggested Facebook is a collection of overlapping, tightly knit, social circles of people who tend to share, discuss and play with each other (Wilson et al 2009).

Within two months of starting social media activity, the decision was made to cease updating the Trove Facebook account, in order to focus on the other social media mechanisms. While this was primarily affected by the technical difficulties above, and the lack of resources to publicise the feed, it may reflect broader differences between social media mechanisms.
With hindsight, it seems likely that Facebook, a service built on a high level of interaction between strongly connected peers, may not be well suited to broadcast-oriented activity, without more interactive content. Some of the most popular library Facebook accounts employ tactics such as: posting questions and polls; integrating with services such as FourSquare; providing mini-games; and opening albums for users to contribute pictures of libraries.

The whole-of-library Facebook page, in contrast to the Trove experiment, enjoys a solid profile and contains content that is more varied and dynamic (YouTube videos, Podcasts, SlideShare presentations), interactive (polls and surveys), and routinely contains calls for action – requesting user assistance with collecting and information seeking. The National Library Facebook page also enjoys frequent opportunities for shared promotional opportunities with external stakeholders, for example, a post about a book launch will be cross-promoted by the author, publisher and members of their networks.

Learning 2: Twitter is an excellent broadcast mechanism

Twitter’s broad distribution of networks leads to much broader dissemination of tweets than Facebook posts, and one that is extremely difficult to track. Because the major distribution of tweets involves networks forwarding tweets to each other, the number of followers an account has does not, in fact, correlate to the number of people seeing their tweets, as counter-intuitive as this sounds. A single tweet can reach tens of thousands of people, simply by being on the right topic, or picked up by the right people (Java et al 2007).

These statistics point to a simple truth: Twitter, while highly interactive, is also more broadcast-oriented than most social media, and easily adapted to distribution of content.

Twitter is a broadcast forum that is frequently used to keep tabs on news, not simply to interact with friends, and hence easier for organisations to leverage for promotional purposes than pure social networking sites. Every day, for each tweet issued, five searches are run in Twitter to create a custom feed (Twitter 2011). Most people engage with Twitter as a source of information, not in an interactive way. Feeds move very quickly in Twitter, spreading information extremely quickly, and also ensuring that a couple of hours later, the ‘Twitterverse’ has moved on.

A vast array of applications exists to view Twitter through a wide variety of platforms. These range from those that will show you five or six streams running at once on a wide screen, down to small optimised-for-your-phone interfaces.

An important feature of Twitter is “retweets”, the mechanism by which a tweeter forwards a message from someone else to his or her own followers. This is generally the way that a particular message goes quickly viral, as tweeters circulate it rapidly from account to account.

A second consequence is that the extent of engagement a Twitter account has is almost certainly a better measure of influence than the number of followers.

At the same time, Twitter is an equal, interactive space. It provides a place to engage with Trove’s supporters on a peer-to-peer basis, which is appreciated by users who view themselves as stakeholders and owners of this public service.

From inception, the Trove Twitter feed had a high level of interaction, both of conversation with users, and also forwarding of tweets discussing Trove. The biggest gains for Trove have been in increased visibility for the service, and in realising the opportunity to interact with Trove fans. The Trove Twitter feed generates a healthy amount of compliments for the service – more than are received via other mechanisms, and has also provided a mechanism to engage with passionate users of the service, in a relatively resource efficient way. The noticeable engagement of media outlets with the Twitter feed is one of the most significant aspects of the service, with tweets regularly generating coverage in new and traditional media outlets.

Results from the trial also indicate that circulating help information via Twitter is popular with users. One of the surprising effects was the popularity of tweets about system outages, new Trove content and updates to the Trove software. In addition, Twitter proved an excellent location to circulate tips, tricks and help about Trove.

Trove has a substantial number of followers on Twitter, with an average increase of 60 followers per month. This includes a diverse range of individuals and organisations. An analysis of Trove’s Twitter followers conducted in July 2011 found that 75% of followers were outside the library community, with a significant proportion of followers identifying as genealogists (7%); teachers (6%); authors (6%) and historians, both academic (4%) and recreational (3%). A diverse range of organisations also follow the feed, including media organisations (6%). The main interests mentioned by Trove’s Twitter followers in profiles were history; government; literature; photography and science.

The @TroveAustralia Twitter account proved a strong success, harnessing the passion and affection of Trove’s volunteer army to increase the visibility of Trove, bring in new users and generate discussion about the many treasures found within the service.

Learning 3: The more the merrier

The Trove Twitter feed, and the regular topic-based forum posts, are run on a distributed model, with centralised coordinators and around a dozen tweeters from around the Library, not just staff working directly on the Trove Support Team.

Having a large number of staff involved in creating tweets and forum posts has a number of benefits:

  • Spreading the work around allows it to be allocated more flexibly. Staff can tweet more when available and inspired, and less when busy and/or uninspired.
  • It ensures a broader variety of Trove resources is showcased. For example, the staff working directly on Trove support during the trial displayed an unusually low level of interest in cricket. Drawing a broader pool of Library staff into participation ensures cricketing topics are still covered.
  • It spreads expertise in using, and managing, social media services more broadly.

At the same time, centralised co-ordination is extremely important. The two coordinators were responsible for:

  • Monitoring comments about Trove and responses to @TroveAustralia tweets;
  • Managing the Twitter roster;
  • Administration (statistics and archiving);
  • Liaising with Library marketing and media officers and the Web Publishing branch; and
  • Ensuring quality control (monitoring, training and advice).

After the first few months of the trial, it became clear that the activities needed two coordinators, to ensure that the monitoring was able to be done daily; and to promote consistency in service, support and coordination regardless of staffing arrangements. Given the reputation risk inherent in social media, it proved essential to have several staff monitoring chatter about Trove. This was highlighted in July 2011, when concerns were voiced via social media channels regarding National Library and Library of Congress cataloguing of WikiLeaks-related material. The Library was able to interact with the concerned community and quickly address the issues raised, resulting in a positive conversation about the Library.

Like all things, engaging with social media requires real resourcing. A distributed model saves resourcing at the content-producing level, but requires significant resourcing at the coordination level, which also requires some management skills. Based on initial time tracking, supporting the regular Trove Twitter feed and a single weekly topic-based forum post requires around 0.4 full-time-equivalent staff, and the majority of the time spent represents coordination work. It quickly became clear, in particular, that producing regular forum posts of length was resource-intensive in comparison with tweeting, which reached a wider audience and generated more discussion. At the same time, the experience of providing more in-depth exploration of Trove as a research tool had some benefits. The team is currently assessing this area.

A surprise benefit from diversifying staff involved in producing feed content was the increased understanding of the service among staff using it as end-users more regularly (often searching on subjects of personal interest), and also the morale boost from staff engaging in the generally positive atmosphere of Twitter discussion about Trove.

Learning 4: Trusting your staff is part of risk management

One of the strongest concerns of any organisation moving into social media space, especially one reliant on a large number of participants, is managing the risk of inappropriate messaging.

All staff participating in the Trove Twitter and blog project are required to read the National Library’s Social Media guidelines and the Public Service Code of Conduct. In addition to the Trove Forum and Twitter aims & outcomes (quoted above), Norms of Practice were also developed. A key aim of these documents is to provide guidelines that enable staff to feel confident being cheeky, current and amusing without breaching the ethical boundaries set by the Public Service Code of Conduct.

A system of tiered permissions works to ensure staff new to tweeting on behalf of Trove are able to develop some experience in formulating tweets in a low-pressure environment.
Two incidents illustrate how this has been controlled:

  • Incident 1: Tweet Sent To Wrong Account

In May 2011, one of the staff involved in regular tweeting, sent a tweet on behalf of @TroveAustralia by accident. The tweet was intended to be sent to the staff member’s personal account, but the @TroveAustralia account was selected instead. The staff member realised and rang the Social Media coordinator, who immediately deleted the tweet, as it was out-of-scope for the Trove Twitter feed. The tweet was visible for less than a minute.

  • Incident 2: Tweet sent from unknown source

In August, a tweet appeared on behalf of the @TroveAustralia Twitter feed from an unknown source. The tweet was a joke about libraries and a current riot in Britain. It was clearly out of scope for Trove, and closer to a political opinion than Trove’s guidelines permitted. The tweet was noticed by the Social Media coordinator within 10 minutes of appearing, and deleted. As a result, the account password was changed, and the permissions for all participating staff were revoked until the new password was circulated to current tweeters.

Measuring impact and deciding if it is worth it

Due to the diffuse nature of social media, it is difficult to come up with accurate measures of impact in cost-efficient ways. When it became apparent that measuring impact was consuming more staff resource than engaging in the activities, a decision was made to simplify this measurement considerably. A similar situation is present with the archiving and preservation of a record of activity.

The resources needed to run the Twitter feed have proved manageable for now. A number of benefits to the team have been generated by the work, including increased usage and familiarity with the service; a closer understanding of the needs, and passions, of Trove’s users; and a boost in staff morale from the regular contact with positive feedback.

Over the course of the trial, many improvements have been made to the workflows of the Twitter feed. Lessons include:

  • Having tightly defined aims and outputs for the Twitter feed ensures a consistent tone and a clear understanding of users as to what to expect when they follow;
  • A thorough understanding of the Twitter service is essential to getting tone and format right;
  • Co-coordinators of the Twitter feed are necessary to ensure timely responses, and to maintain broad participation;
  • Broad participation in tweeting and blogging provides diversity of content; increases the staff ownership of Trove; and increases experience with key social media principles;
  • Having staff running a service also managing a social media channel both benefits the team, and ensures timely and accurate content;
  • Risk is best managed through training and close communication, rather than central moderation; and
  • A mix of rostered and spontaneous tweets maintains a relevant feed with a minimal disruption to daily workflows.

In contrast to the Trove Twitter feed, the Facebook experiment failed to generate benefits. This may be because Facebook is less suited to the broadcast of information, as this space is much more focused on interactive social chat, and playfulness, generally within closed personal communities. It may be that Trove users are less interested in engaged with Trove on Facebook. Future attempts to build Facebook applications should take this into account.

The blog posts, while not as successful as hoped in drawing new readers in, remain useful.
Currently, statistics collection and reporting; along with archiving of tweets, comprises a significant workload for managing the feed. This is also the case in managing the National Library’s corporate Twitter feed. Software to simplify rostering, statistics and archiving may prove cost-efficient, and deliver long term benefits as activity in this area is increased. This is currently under investigation.

Future trends

Trialling social media marketing for Trove was a valuable learning experience for the National Library of Australia. As this paper outlines, much of what we tried did not work, and often the aspects that worked, did so in an unanticipated way. Some of our anticipated terrors did not eventuate; many we were able to moderate by remaining flexible. Flexibility also proved to be the key to our successes – being able to build on what worked well, and quickly abandon what did not. Perhaps the greatest challenge that remains is that of resourcing.

The Library, like most cultural institutions that were quick to experiment with social media, quickly learned that it really does mimic the ‘free puppy’ scenario – it might not cost anything up front but as time passes the hidden costs, particularly to your time, become increasingly visible. Accordingly, an important part of the learning process for cultural institutions is around improving our ability to define aims and audiences and to assess the resourcing requirements. For example, once a feed picks up passionate supporters, there is a significant cost in support and goodwill to ceasing the activity.

As our understanding grows further, we will be presented with the opportunity to uncover and engage more with micro communities that house enthusiastic, passionate and keen users and potential users. While there currently remains a strong impetus and motivation for building institutional or ‘corporate’ brands in the social media space, it is possible that in the future institutions will move further away from universal corporate brand and towards the style and tone of the community that each channel serves.

This approach has already been utilised with great success by the ABC, who have uniquely branded both their online communities and mobile apps to reflect the target markets of each.
The success of the Trove social media pilot demonstrates the potential of niche and service-based communications for cultural institutions, while still underscoring the importance of tailoring an organisation's strategy to particular services.

Cultural institutions should be giving consideration to the possibility that niche communities will become the norm and the foundation of social media activity. This raises a number of questions we need to consider, such as what will this mean for us in terms of online engagement? What are the implications for us if we want to move collection content from existing service delivery platforms out to where the people are engaging in activities that directly relate to our collections? And, as the resourcing demand grows, can we afford to maintain traditional media channels, while establishing a variety of specialised social media channels? Or, will we be prepared to resource building specific communities around distinct services at a cost to our whole-of-organisation communication to everyone?


List of cited online services & social media

National Library of Australia

Australia Dancing 
The eloquent page 
NLA Facebook
NLA Flickr
NLA SlideShare
NLA Twitter Page
NLA YouTube
Picture Australia
Trove Twitter Page

Other organisations

British Library blogs
Kew Gardens Flickr
Library of Congress blog
Lorcan Dempsey Twitter
Mark Scott Twitter
NYPL blogs
OCLC WorldCat
Penn Tags
Plateau People’s Portal
Roy Tennant Twitter 


Facebook, 2011. Statistics. Accessed November 25, 2011.

Fishkin, Rand, 2011. Facebook, Twitter influences Google search rankings.

Holley, Rose, 2009. Many Hands Make Light Work: Public Collaborative OCR Text Correction in Australian Historic Newspapers NLA Staff Papers.

Huberman, B.A., Romero D. & Wu, F., 2009. Social networks that matter: Twitter under the microscope. First Monday 14 (1).

Java, A, Song, X, Finin, T & Tseng, B., 2007. Why we Twitter: understanding microblogging usage and communities.

Kaplan, A.M. & Haenlein, M. 2009. Users of the world, unite! The challenges and opportunities of Social Media. Business Horizons 53 (1) pp. 59-68

Kwak, H, Lee, C, Park H & Moon S., 2010. What is Twitter, a social network or a news media? 19th International World Wide Web Conference, Raleigh, North Carolina.

Neilsen, 2011. State of the Media: The Social Media Report Q3 2011.

SocialMediaNews, 2011. Social Media Statistics Australia – October 2011.

Twitter, 2011. The engineering of Twitter’s new search experience.

Wilson, Christo; Boe, Bryce; Sala, Alessandra; Puttaswamy, Krishna; & Zhao, Ben, 2009. User Interactions in Social Networks and their Implications.