• Français
• Home
• Artefacts Canada
• Funding
• Conferences & Training
• Jobs
• Members

Social Computing from the McCord Museum's Website

Introduction

Montreal’s McCord Museum obtained funding from the City of Montreal in 1996 to create the very first version of its website. The latter was launched in 1997, and contained a database of over 500 artefacts, representing the ‘highlights’ of the McCord holdings. Various grant opportunities have since allowed it to expand to contain over 135,000 documented images (including 80,000 from the Notman Collection), 80 Web tours and video clips, some 120 thematic files, 36 games, learning resources for 49 different subjects, and numerous features that allow virtual visitors to use, contribute and share information. It now receives close to 1.6 million visits from nearly 900,000 unique visitors.

This paper focuses on the implementation of various social computing features on the McCord Museum’s website. It reviews the added features, the justification for adding them, the management process used, perceived risks, financing, ongoing maintenance, and the impact, to-date, of these features.

Background

In 2000, McCord added user account capabilities to its website which allowed its users to create Web tours. From that point on, users were able to add text to their own selection of McCord images, and to then share this information with friends. Shortly thereafter, a more formal user account system was put in place, allowing users to order photographic reproductions of images viewable through the website. Other features were subsequently added, including the ability to add a user’s own personal images to Web tours, and many of these early components became the building blocks for future work, as is the case for the social computing features that followed.

Throughout 2007, McCord implemented a formal ‘My McCord’ section with new functionalities on its site, including the following social computing features:

1. A tag box that allows registered users to add a publicly viewable short descriptor to any McCord image.
2. A tagging game that allows users to play against one another (or the computer), by guessing what tags the player’s counterpart is entering for the same McCord image. The goal is to identify a common tag for as many images as possible, within the allotted time, thereby populating museum images with tags.
3. A feature that allows registered users to add comments to McCord images that can be made viewable to the public.
4. A feature that allows registered users to share their Web tours with the general public instantly, without intervention by the McCord.
5. A feature that allows registered users to annotate images directly within the image.

This work is ongoing, and in January 2008, McCord modified an existing image pairing feature so that paired images could be shared with others.

What objectives is McCord trying to achieve by implementing these social computing features?

McCord views its website as a pillar of the museum itself. The website helps engage those who visit the physical museum, and those who otherwise would not, or cannot. According to Nicole Vallières, Director of Research, Collections and Programs at the McCord Museum¹, culture must be shared, and thus, the public must be invited to interact with the collections, with the interpreters of these collections, as well as with one another. Accordingly, McCord is adding Web 2.0 features that lend themselves well to helping Canadians share their culture with each other and the world.

How is the audience identified?

McCord uses both Google Analytics and Web Trends to analyse aggregate visitor behaviour. McCord also studies traffic within its physical ‘walls’, and remains abreast of existing demographic research, such as the work on museum website traffic conducted by Paul Marty, Assistant Professor, at Florida State University.² Accordingly, McCord knows that its visitors are adults interested in history, tourists with an interest in First Nations content, teachers and students who visit the museum on field trips, specialized audiences (with interests in photography, material culture or archival documents), and researchers.

Site features are designed with this audience in mind. In addition, McCord is using social computing sites to reach an audience that is underrepresented within the usual scope of museum visits. It has uploaded material to YouTube to promote and give access to its content. The museum will soon follow up with uploads to Flickr.

Who builds and maintains the website?

Until October 2007, development and general management of the website were priorities of the Collections and Knowledge Management Department, from which a core team of four persons was responsible for both the site and site-related matters. Apart from the Department’s Director who holds a PhD in ethno-history, team members included an information coordinator, a webmaster-programmer with a detailed understanding of all technology used on the site (and the ability to implement changes), and an Online educational program coordinator, who holds a PhD in education. Last November, the McCord created the Multimedia Programs Department, one of six departments under the directorate of “Collections Research & Programs” that now answer to the team’s original director.   This new Department has a similar size and structure to that of the original team, but it is now dedicated exclusively to the website and all future Web development aspects.

This team maintains the site and designs all Web features. It builds the smaller components in-house, and outsources larger tasks. However, no outsourced work is ever turn-key. The webmaster, who is also in charge of Web architecture, is ultimately responsible for any technology found on the site. This point is important – the McCord team has full control of its own website, and understands how it works. McCord’s proprietary content management system, for instance, was built in-house as the website grew, and this allows the museum to ensure a stable and reliable environment on which to build future components.

How are new features added?

Whenever there is a need to add a new feature to the website, the core Web team discusses the issue and submits its recommendations to the Department's Director. Although the addition of website features is usually initiated by team members, any staff member may propose site features at any time. The value of each feature is then weighed for its ability to serve the Museum’s objectives. In any case, a new feature is always considered as an experiment. Many of the site’s features have not been in the field long enough for the McCord staff to determine whether or not the audience will make use of them or find them useful.

If a proposed feature is exceptionally expensive or complex, a prototype is first built and tested with a focus group that may consist of an informal collection of staff or family members, or a more formal audience, including primary and secondary school students. Testing is carried out throughout the development process using a mirror site not available to the public. Once a feature is launched, its use is closely monitored. If a new feature is found to have a positive impact, it is kept and further integrated into the site. It may also be modified and retested, or removed from the site altogether. In a way, the McCord website is in a perpetual beta phase, as Web 2.0 dictates that constant improvements must be made and that content must be kept up to date.

What costs are involved and how is it paid for?

It is estimated that approximately $7. 5M have been invested to date in building and maintaining the McCord website³, $5.5M of which comes from external sources.

The following are other numbers that have been determined for specific social computing features:

• Cost of the “My McCord” personal Web space – approximately $100,000 in external contracts, including:
  • $25,000 to develop background functionality (such as file and image management);
  • $75,000 for account management features and the adaptation of existing features for use in the “My McCord” space.
• $50,000 in external contracts for the image tagging game.
• An internal programmer developed a search engine and modified it through its many iterations.
• Other internal improvements dealing with the development of these features include management activities, coordination of pilot (test) projects that involve community members (such as teachers and students), and cooperating with partners. All of this work is ongoing, as it has been for the past decade.

McCord is fully aware that by using an incremental approach to website building, it is increasing development costs, since existing codes often need to be modified to accommodate new features. There are, however, certain advantages. For example, the museum can take whatever time is necessary to test and modify features, and adapt to new technologies and trends as they emerge. It can also spread development costs over multiple fiscal years.

Two main sources are responsible for funding the site. First and foremost are internal operational funds. McCord sees its site as the pillar of its institution. Consequently, its budget reflects this fact. Operational funds cover the salaries of employees who work in the Multimedia Programs Department, and are also used to finance part of the outsourcing of larger projects.

McCord also benefits from external funding sources, and the museum has had ample success in partnering with other organizations to conduct website research and development. For instance, current project partnerships include McGill, Laval, Sherbrooke and UQAM universities, schools and educational boards, the Canadian Heritage Information Network (CHIN), and the National Research Council (NRC). For this very reason, the success it has known in the past, and an understanding of how it fits in with the objectives of various funding programs, McCord has been competitive when applying for project funding. These funds allow McCord to develop and test new features that it could not otherwise afford. However, no feature is ever developed (regardless of where the funding comes from) unless operational funds are in place to maintain it in the long-run.

What risks were perceived prior to implementing these features?

Five major risks were identified prior to the implementation of these features. They are as follows:

1. Technical (the ability to implement and manage features) – Minimal. McCord has the necessary in-house technical expertise as well as an in-depth knowledge of its own website. Social computing features are technologically stable.
2. Financial (the cost of implementing and maintaining features) – Minimal. External funds were used to implement these features, and current staff should be capable of undertaking the related maintenance workload.
3. Human Resources (managerial and technical skills required) – Minimal. In-house skills are strong.
4. Publication of results (adoption of the features) – It may take as much as three years to have sufficient data to determine the impact of any given feature.
5. Legal / Political (liability, copyright, privacy of data, privacy of minors, etc…) – McCord controls all aspects of image copyrights. It doesn’t need to obtain further permissions to publish content on its own website. In turn, institutions and individuals who have been granted a Creative Commons license are allowed to use this material for non-commercial purposes.

   Beyond this, all other risks, such as those on the legal and political front surrounding Web 2.0 features, are the same as all those perceived by other institutions facing similar decisions. These concerns are addressed by trying to be as aware and as responsible as possible, without being too restrictive. The Museum strives to remain abreast of current legislation in these matters. According to Dr Vallières: “The biggest risk perceived by McCord is failing to get people involved.

6. Other Risks:

• Having social content mistaken for authoritative content :Such an instance occurred last year when deep links from a Google Web search landed users in the centre of a student’s “My McCord” space, where a homework assignment had been published, complete with errors. The solution was to provide better orientation on all pages so that users coming into a deep link from elsewhere would understand the context of what they were seeing; for example, colour coding, user contributions and adding a disclosure line.
• Risk of success: If a feature works too well, it can also present a problem as one needs to have the required human resources in place to manage these results – a situation described by McCord as “growing challenges”. This risk has in fact become obvious to a lesser extent; and additional work has been required of existing staff to correct typos and inaccurate or incomplete McCord content identified through user feedback (mostly identification of places within 19th century photographs). However, this “problem” is seen as a benefit in that McCord’s content is being vetted by its users.

What is the cost of maintaining these features?

Maintenance of the features themselves is negligible, although in-house modifications will constantly be made as these features continue to be tested and improved.

The moderation of social content has also been minimal. The McCord site has a self-moderation feature that allows users to report inappropriate content, and beyond this, comments, tags and other social content can be sorted by McCord staff in reverse order, according to the date that content was added or modified. Up until now, no abuse has been brought to the museum's attention. A generic email capability will soon be implemented, and this will automatically reply to users submitting feedback so that they can know immediately that their comments were received.

Maintenance costs played a role in the decision not to include a blogging feature. Originally, McCord considered implementing a blog to allow curators to share their work with the public, but the time required to maintain the blog was deemed to be too much of a burden for current staff. A well-known fact is that if a blog is not kept current, the site quickly loses its visitors’ interest.

Results

As less than one year’s worth of complete traffic information is available, the data is inconclusive. Since McCord does not promote its launches, it can take up to three years before the true impact of a feature is known. The following statistics may, however, be of interest

1. Since its introduction on August 9th, 2007, over 215,000 tags have been collected through the tagging game. During this nine-month period, 6,764 games were played.
2. Between July 15th, 2007 and May 15th, 2008, some 2,000 Web tours were created by 1,180 different users.
3. Roughly 6% of all searches conducted on the site are searches that use the tagging feature.
4. Seasonally adjusted visits to the overall site have increased roughly by 10% (over the previous year’s data) since the soft launch of the redesigned website in July 2007. Virtually all of the increase can be accounted for by repeat visitors – this is in keeping with regular site growth – and regular site development.
5. An average of 104 users per month have registered for the new McCord user accounts – more so during winter months (in keeping with overall traffic on the site), for a total of 1,410 accounts as of May 15th, 2008 (10 months after the implementation of this new feature).
6. Each registrant has logged on twice (on average).

Further to the measurement of web traffic volume, other factors will also be measured to determine the degree to which visitors interact with the site, including such metrics as time spent online, documents shared, emails received, images tagged and content shared.

User engagement is the ultimate goal of social computing features. However, when traffic increases, history has shown that this translates into increased museum attendance,⁴ and enhanced gate revenues are an added bonus!

Summary

While it is still too early to determine what lessons can be learned from the addition of these features to the site, it is possible to learn quite a bit from the methodology used by the McCord Museum. The latter sees its website as its pillar and uses it to help meet its objectives. However, concerns regarding abuse of Web 2.0 features have not been justified thus far and, when the latter is implemented correctly, the moderation of social computing features takes very little time. Although the museum had to deploy additional efforts to respond to an increase in feedback on the inconsistencies found in existing records, this increase was seen as an engagement on the users’ part, and the public vetting McCord’s data was certainly perceived as a benefit.

As for funding, McCord finances its site by using external funds to develop larger and more expensive components. It uses internal core funds to develop and experiment with smaller features. No feature is ever developed unless core funds are in place to sustain its subsequent operation.

Finally, McCord’s technical staff possesses a strong understanding of its website, and it can therefore experiment at will. While the results of newly-developed Web 2.0 features are not yet clear, McCord will continue to proceed as it always has by monitoring user response to new components, modifying these as necessary, and learning what works best.

For more information, visit the McCord Museum’s website at: http://www.mccord-museum.qc.ca/en/, et la version française à : http://www.musee-mccord.qc.ca/fr/

1. Current title held since October 2007. Previously, Director, Collections and KnowledgeManagement, McCord Museum.

2. Paul Marty, ‘Museum Websites and Museum Visitors: Digital Museum Resourcesand their Uses’, Museum Management and Curatorship, March 2008, Volume 23, Number 1, March 2008, pp. 81-99.

3. Including staff time - Interview with Nicole Vallières, March 14, 2008.

4.http://www.pro.rcip-chin.gc.ca/sommaire-summary/2004survey-2004survey-eng.jsp