Social Media Metrics
By David Stuart
Social media has been adopted by every type of library in recent years, from small special libraries to large national ones. Many now host blogs and wikis, are members of numerous social network sites, and even participate in virtual worlds. These sites and technologies offer new ways for library staff and users to communicate and collaborate. However, with so many different technologies and sites available—and with more emerging all the time—it is important that librarians develop methods for measuring the use and effectiveness of the technologies so that time is not wasted and the implementations are justifiable to upper management.
Social media—the most popular are blogs, wikis, social network sites, and microblogging—offer both opportunities and difficulties in the establishment of metrics: Social media adds a level of qualitative information to the quantitative data traditionally made available through web analytics. However, the quantitative information is often restricted and not easily comparable among sites.
Blogs are regularly updated websites that display posts in reverse chronological order and are one of the more established social media technologies. These are regularly used in libraries both to share general news about the library and to provide subject-specific information.
In terms of social media metrics, blogs have the big advantage of allowing the use of traditional web analytics. Whether an institution decides to host its own blog or makes use of the hosting facilities provided by companies such as WordPress and Blogger, a small piece of code can be automatically added to each page so that information can be captured about visitor behavior. In addition, there are a number of other metrics, not applicable to traditional websites, that can provide a better indication of the usefulness of the blog.
• Blog comments. Comments are one of the most important features of blogs, allowing visitors to share opinions. While a large number of page visits may be the result of high search engine ranking of an organization’s site, the number of comments shows engagement with the audience and can be more important.
• Blog reactions. Although people blog individually, the blogosphere can be seen as a community, with lots of blog posts responding to other people’s posts. Although certain blogging platforms allow these reactions to be highlighted through “trackbacks,” where these are unavailable, you can discover them through blog search engines such as Technorati.
• Blogrolls. A blogroll is a list of recommended blogs that is often included in a blog’s sidebar. Think of this as more important than a single link from a blog post to another blog post. These links are generally not found by blog search engines. But they can, at least currently, be found among all the incoming links identifiable through the link domain operator in Yahoo! Search. For example, entering linkdomain:www.onlinemag.net will retrieve a list of all the in-linking pages to ONLINE’s website.
• RSS readers. Subscribing to an RSS feed provides a way for users to discover whether or not a blog has been updated without having to visit the site. Feed management providers such as Google’s FeedBurner enable a detailed breakdown of the number of subscribers and where they are based. It is also possible to get an indication of the number of subscribers from Google RSS Reader. Although this will be only a proportion of the total number of subscribers, it enables a comparison between web feeds. Google Reader is increasingly becoming the feed reader of choice.
Each of these metrics shows a greater level of engagement with the blog than would merely be shown by user traffic; these are not users who stumble across a site accidentally but those who are interested in what is being said. The importance that is ascribed to one metric over another depends on the purpose of the blog. If the purpose of a library’s blog is to get feedback on the services that it offers, it would consider the comments a far more relevant indicator than page views, whereas if the purpose is merely highlighting the latest library acquisitions, then the number of page views may be considered most important.
Wiki software can be used for the collaborative creation of webpages with a WYSIWYG (what you see is what you get) interface. Wikis have been used within libraries for the crowdsourcing of subject and local information guides.
The success of a wiki may be quantitatively measured in several different ways—number of pages created, number of editors, and the amount of edits.
Once again, the importance ascribed to each of these metrics will relate to the purpose of the social media. If a library is trying to create a comprehensive local information guide, the number of pages that are created are of primary importance. Although it may be desirable that a wide range of people contribute, it is not a necessity. However, if a library uses a wiki for a local history project, trying to capture the personal experiences of the local community, then the number of editors is likely to be considered far more important than the number of pages created. The amount of page edits provides an indicator of interest in a topic, or its contentiousness.
Social network sites
The term “social network site” can be used to refer to a multitude of different sites that enable users to create a profile page and connect to other users. Social network sites may be primarily people–focused—such as Facebook or LinkedIn—or content–focused—such as SlideShare, You-
Tube, or Flickr. Libraries use many of these sites to share content and to connect with users. Some of the platforms provided by sites such as Facebook have allowed libraries to create applications that enable users to interact with library systems without having to visit a library’s main website. For example, a number of libraries have Facebook applications that allow users to search the libraries’ OPACs.
Social network site metrics are heavily dependent on the information that a site decides to share. This can vary considerably not only from site to site but also according to a user’s type of account. For example, Flickr offers premium users detailed information on viewer numbers and referring sites, whereas nonpremium members are merely provided with the number of views of each photo. In comparison, YouTube offers detailed information on viewers of videos to every user, even providing demographic information on the viewers.
The one piece of information generally provided by all social network sites is the number of other users in the network. This may be the number of subscribers to a YouTube channel, the number of members of a LinkedIn group, or the number of followers in SlideShare. While this may provide an indication of initial interest, it is not necessarily an indication of active interest. For example, although a Facebook or LinkedIn group may attract numerous members, membership is not particularly intrusive and may be continued without active participation. It is, therefore, appropriate to pay more attention to the number of comments, which can also be left on social network sites.
It is also important to recognize that as social networking sites upgrade their offerings and new sites launch, users emigrate from one site to another. It is important to keep track of the popularity of the sites being used and to watch for the emergence of sites proving popular with users. Web traffic sites such as www.alexa.com can provide information on the changing fortunes of websites, providing detailed information on user demographics.
Microblogging generally refers to the practice of blogging with 140 characters or less, although many applications built around the microblogging sites also allow for the uploading of pictures and short audio and video content. Despite the name, most microblogs have more in common with social network sites than blogs: The focus is generally on the network of friends and followers, with microblogging happening on a few large websites rather than taking place on a user’s personal sites. Like blogs, microblogs can be used to share general news about the library and subject-specific information. However, they can also be used for asynchronous conversations with librarians responding to users’ queries. Twitter is the most popular of these sites by far, with numerous news stories and celebrity endorsements leading to exponential growth during the first half of 2009. Others include Plurk and Yammer.
Twitter itself offers little in the way of statistics beyond a user’s number of friends (the Twitter feeds a user follows), followers (the number of people who follow a user’s feed), and a user’s number of updates. Although the number of followers may seem to be the most appropriate indicator of a feed’s success, it should be taken with a pinch of salt. The number of followers is highly influenced by the number of feeds a Twitter user follows, with many people automatically reciprocating the following of a Twitter stream if the user is not obviously spam. Equally, there are a large number of accounts that follow Twitter streams in the hope of gaining a reciprocal following rather than having any interest in the content of the stream.
On the other hand, you can argue that the lack of a large number of followers is not necessarily an indication of lack of use. The strength of microblogging is in sharing information across the network, beyond a Twitter stream’s immediate followers. If followers find a particular update useful or think it may be of interest to their followers, they can copy the update and “retweet” it in their own streams. As such, a handful of influential followers can potentially share a person’s updates with many thousands of users who are not following that person’s stream.
It is therefore important to refrain from taking the number of followers at face value; instead, take into consideration how the followers interact with a Twitter stream—whether other microbloggers are addressing updates to the account or retweeting the updates.
There are a number of additional sources that can form the basis of microblogging metrics:
• TweetStats (www.tweetstats.com) and other sites based around Twitter data can provide additional insights into how a Twitter feed is being used and the network of followers.
• URL shortening services are particularly useful for highlighting external resources when brevity is paramount. Some of these services, such as http://bit.ly, provide statistics about how often these links are being followed.
• Twitter Application Programming Interface (API) provides a method of automatically collecting the Twitter data the staff are most interested in; it is a particularly useful metric if library staff are investing a lot of time in communicating on Twitter and if staff with programming skills are available. The API can be based on the network of Twitter followers or the content of their updates.
Useful social media metrics
As can be seen from this brief overview of some of the social media technologies used in libraries and the potential quantitative and qualitative data that can be collected from them, there are no simple solutions to social media metrics. Are 1,000 blog page views worth more or less than 100 retweets? Are 100 YouTube subscribers better than 200 LinkedIn group members? How much attention should be paid to the quality of interactions in comparison to the quantity? How many short Twitter replies are worth one long blog comment? The answers to these questions start with the reasons for the adoption of the particular social media.
• Have an explicit purpose for a specific social media. The true potential of social media in libraries comes from allowing staff members to experiment with the application of the different technologies. Nonetheless, to prevent time being wasted with little purpose on a technology merely because it is “the next big thing,” it is necessary that staff members are available to define the purpose of the technology at the outset. Purposes can change over time as staff members respond to user needs. However, unless there is an explicit purpose, it is impossible to construct useful metrics.
• Identify relevant metrics. Based on the explicit purpose, the initiator should be able to identify potential methods of measuring the desired outcomes of a specific social media. Such metrics will change with any changes in the purpose of the social media.
• Set realistic benchmarks. While return on investment depends on the time that is spent with a particular type of social media and the perceived value of certain metrics, it is nonetheless possible to determine whether expectations are realistic by comparison with those statistics that are publicly available for similar institutions. Having more Twitter followers than Oprah Winfrey’s millions is not only unrealistic, it is undesirable since responding effectively to such a large community would take effort that could be more effectively focused on a specific group of users.
• Regularly reappraise the situation. The purpose of the technology, the appropriate metrics, and the appropriate benchmarks can all change over time. It is therefore important that each of these three steps is regularly reappraised.
As libraries become more involved with social media, implementing blogs, wikis, social networking, and microblogging, the measurement of the impact of these technologies to assess their effectiveness and return on investment is essential to justify their adoption.