Macquarie University is an innovative university located on a single campus in Sydney, Australia. Macquarie was founded in 1964 as a second-generation Australian university with a focus on interdisciplinary research and teaching in the sciences, social sciences and humanities.

In line with Macquarie’s reputation as one of Australia’s most innovative institutions, the Macquarie University Library (MUL) plays a key role in providing information resources to the University community.

Delivery of electronic services has been a growing focus of all libraries over the last ten to fifteen years, and recently there has been a sharp increase in the priority given to online services. The core element in this strategy for MUL is its website, which allows authorised users to access most of the material in the Library’s collections.

The website consists not only of a public brochureware site, but also the ‘business end’ information portal; a powerful set of tools that allow users to search out and retrieve books, academic journals and other material.

However, there was a general recognition throughout MUL that the website needed improvement, both in its appearance and usability, but also in its usefulness for the intended audience.

MUL also recognised that it needed to understand its audience better, so it could deliver appropriate targeted services both online and in person.

MUL sought the assistance of Step Two Designs to begin the process of evaluating and redesigning the website. This case study aims to give an overview of the process undertaken and the key outcomes.

Project goals

Library staff had seen the website through several previous redesigns, so they had a good understanding of what is involved in designing and maintaining a website, but they needed some assistance in specific areas.

In particular, we helped with user research, information architecture and general web design and usability expertise.

The goal was to review the website in preparation for a major redesign. This required substantial user research, and it was decided to undertake this more broadly, because gaining a clear picture of MUL clients and their needs would also be useful for the website.

It was also necessary to gain an understanding of the organisation itself, how it functioned and how this affects the delivery of services to Library clients.

Based on the research, a set of strategic recommendations and a roadmap for future website development were formulated.

However, the focus of this article is on the development of a new audience model, including personas.

Review the website in preparation for a major redesign

Overview of methodology

Ultimately, the MUL website will have many aspects redesigned based on this work, but the immediate product of this project was an evaluation of the current site and the creation of an audience model. These will feed into information architecture design and implementation work further down the track.

The project was made up of two main activities: an expert evaluation and the user reseach.

The expert evaluation took a long hard look at the actual website and identified a number of areas that needed attention. Such an evaluation covers the practical and heuristic aspects of all websites, irrespective of the specific audience.

However the main focus for the project was on the user research, looking at the website from a strategic viewpoint, understanding its audience and how they use the library.

The focus was on understanding the audience

Understanding clients’ needs

It is very important to base design on a good understanding of requirements. Too often, websites—as well as intranets and software applications—are designed purely on the basis of functional or technical requirements, or the opinions of those who create them.

This was definitely the case for the MUL website. It is a highly functional website, providing several search interfaces as well as access to many external academic databases. In effect the key parts of the website are more akin to software than a normal static website.

Such a site requires carefully crafted user interfaces, that work efficiently, are easy to understand, and possibly most importantly, are useful for users.

A number of user research techniques were employed during this phase of the project, including:

• interviews with a variety of staff
• focus group with key stakeholders
• analysis of web analytics
• interviews with MUL clients
• usability testing
• contextual observation

Good representation of clients was achieved from each of the categories found in the Library’s established view of its clients.

The established view of clients

MUL, like most academic institutions, has a view of its clients, based on traditional academic structures.

The complete list of all possible clients includes:

• Library staff
• University academic staff
• University general staff
• Undergraduate students
• Postgraduate students
• Higher degree research candidates (PhD)
• Commercial entities (eg strategic partnerships with the University)
• School and TAFE students
• Local residents
• Staff from other libraries

This highly hierarchical model lumps disparate groups, such as all undergraduate students, together.

This view is probably useful for many purposes, such as administrative organisation. However, for the purposes of evaluating and redesigning the website, a more detailed model of the audience was required.

One key issue was the inclusion of ‘library staff’ on the list. While library staff do make use of the website—primarily in lieu of an alternative system for them to access library systems—they should not be the target audience group.

Librarians have strongly held views on information delivery, as a result of their long training and experience. This is in stark contrast with the situation of most library clients and website users.

In reality, anything designed from the librarians’ point of view is more than likely not serving clients well.

While there is some argument for ‘expert interfaces’, the only effective method of deciding what should be provided and how it should work, is to understand the needs of the users.

Library staff should not be the target audience group

Audience modelling

The wide variety of clients have different goals, behaviours and attitudes. Upon analysis, natural patterns emerge, suggesting groups within the total audience. Based on these groups, an audience model was created to define the audience.

Fortunately, there was plenty of data available on MUL clients. In addition to the primary research undertaken (eg interviews), a regular library survey undertaken by most Australian University Libraries and the Macquarie University annual report provided a wealth of useful data on students and staff.

The large sample size of these secondary research sources was very useful in terms of gaining more detail but also validating the findings of the qualitative research.

Analysis of the surveys suggests several options for segmenting MUL clients, including: frequency of library use, frequency of online visits, frequency of campus visits, ‘type’ and ‘category’.

These views suit the purposes they were published for, and maintain the established view of clients, but don’t necessarily provide a useful model for website usage.

Instead, a new model was created by looking at how often clients use the library (or how much they need it) and their experience in using the library (or how serious they are about using it).

The resulting segmentation is illustrated in the chart on this page.

Throughout the development of the audience model, the evidence collected about library clients was used to further refine the model.

Natural progression

It was obvious that many Macquarie students are very occasional users of the library.

They have little or no need for scholarly research because their courses do not require it. Meanwhile, other students have to use the library a lot because their courses do require it.

And library use changes over time. Initially, students have very few library and research skills. As they progress further through their academic career, students build up their skills. Demands on them also change, for example moving on from undergraduate study means doing more research.

By the time they reach the level of PhD or academic, clients are more or less professional researchers. They are familiar with library systems and scholarly research is a core part of their careers.
But life experience and circumstances also play a part. For instance, a studious undergraduate student has a very different approach to a less motivated individual. Mature-age students typically strive to ‘go the extra mile’ to do well at University.

Audience personas

For each audience segment, a draft persona was created. A persona is a profile that represents the key characteristics of the clients in that segment. Personas are an excellent way of capturing and communicating the needs, behaviours and attitudes of each audience group. They were chosen for this project in order to focus design efforts on the end users of the site.

The personas for the MUL website are:

Nicole the novice scholar

Nicole (see figure 2) is an undergraduate student whose course requires a fair bit of research in the library. Students like Nicole, studying humanities and similar subjects, need to use a lot of academic material: books, reserve books and journals as well as media and the internet. However, Nicole is not well equipped for this kind of work when she arrives at university and is unfamiliar with information systems or research strategies.

Eric the experienced researcher

Eric (see figure 3) is a lecturer for the University, and like all academic staff at Macquarie he also has his own research projects. Eric and his peers, who include PhD students, are full-time academics. He is trained and experienced at using libraries and the information systems found in them.

Sunita the studious student

Sunita (see figure 4) is a career woman studying part-time to get her Master’s degree, something she takes very seriously. Students at this level, as with mature-age undergraduate students, have more life experience. They know what they want and are studying to get it. Sunita strives to go the extra mile and get the best possible results.

Feng the infrequent user

Feng (see figure 5) does a course that doesn’t require much use of the library. He and other students doing accounting and finance degrees—as well as IT and law—rarely come to the library to do research or use academic material. He uses the library systems so infrequently that he doesn’t get used to them or build up tactics and strategies for success. Feng needs help to use the library.

Sarah the high school student

Sarah (see figure 6) is not at university yet, but she occasionally uses the library. She and other casual users of the library, such as members of the local community or amateur historians, use the library rarely and lack confidence in an academic environment.

Developing the personas

The defining characteristics that set each persona apart from the others, form an essence. It is good to try and capture this essence in the persona’s title.

Fleshing out the personas involves adding details such as a name, photo and demographic details. The figures on page 4 show fragments of longer documents, but represent the core of the personas.

The longer versions add further embellishment in the form of narrative style scenarios of use.

In all cases the details added to the personas were based on data collected through either primary or secondary research. Only details (or attributes) that are relevant to clients interacting with MUL, or to staff understanding how to serve clients, should be included.

Personas were based on data collected through research

Persona prioritisation

It is simply not possible to design something that meets the needs of everyone. A website cannot be designed to be optimal for all of the personas created.

Creating personalised interfaces, or content, for different personas is one solution, but there will be times when decisions need to be made about the choice of design solution.

This is precisely the point of personas, to assist design decisions. To resolve competing or conflicting needs, the personas must be prioritised into a primary persona, and several secondary personas.

For the MUL website, Nicole is recommended as the primary persona. This means that most design decisions will be made to suit Nicole (and the clients she represents).

Nicole was selected as the primary persona because the students she represents are in need of assistance and have high future potential. That is, they don’t have enough library skills to do their best with the existing website, and it makes sense for them to get better at it since their courses use the library a lot.

Feng and Sarah were not chosen as the primary persona because they use the library so little that they will have difficulty using it no matter what. Things can be made easier for them, but there is only so much improvement to be made.

Of course, efforts should be made to accommodate the secondary personas, but Nicole should be given priority. A good example of when this is necessary is the labelling of navigation items. What would Nicole call this?

Other personas

In addition to primary and secondary personas, it is possible to add further personas to the audience model in the future.

This might include the creation of an ‘anti-persona’ (also known as an ‘unimportant’ persona). For the MUL website this could be used to represent library staff.

A common mistake made by website designers is to design for themselves. So an anti-persona would further clarify the boundaries of the website audience, and that the website should not be designed for librarians.

Clients changing hats

The prioritisation of personas is not set in stone. The primary persona may change as library objectives change over time, but of more immediate concern is the need to set different priorities in different sections of the website.

For example, Nicole is the primary persona for the overall website, but within the help section, it might be Feng that is selected as the primary persona. Design decisions within that section would be made according to his needs.

Another special case concerns the general (or non-academic) staff working for the University. In performing their job, they are unlikely to use the library.

But, as it happens, many University general staff also study part-time. In this case, they would assume the most appropriate persona (eg Feng, Nicole or Sunita). Indeed, based on their levels of library expertise, they may assume the Eric persona.

The prioritisation of personas is not set in stone

Evolution of personas

The personas were presented to MUL as draft versions, and for good reason. They are a ‘work in progress’ and a continued, collaborative effort will be required to refine them as time goes on.

As further data is gathered, and the website team become more experienced with the use of personas, improvements can be made.

This approach also recognises that the audience will slowly change, and the personas (and possibly the segmentation) will need to accommodate this change.

For instance, a persona might be split apart in order to give a smaller segment better focus (eg research assistants). Alternatively, two or more personas might be joined together as needs or skills converge in different groups.

Further effort will be required to refine personas as time goes on

Strategic review

The user research provided good information about the MUL audience, and in addition, several key themes emerged about the way clients think of the library and go about using it.

For example, one such theme was ‘the library is just a building’. To the vast majority of clients (ie students) the library is primarily a building. They see it as a place to go to study or take a break, and are largely unaware of what else the library has to offer.

These findings can have a direct impact on the strategic and tactical decisions made by MUL, not only about the website, but also about the way the library operates in general.

Each theme was documented in the final report given to MUL, supported by direct quotes from clients and, where possible, examples to clearly illustrate the situation.

A roadmap for website development

Moving forward with the project involved planning what to do with the website and when to do it.

Improving the usability and effectiveness of any reasonably-sized website requires improvements on many fronts, and attempting to tackle all of these in one go is unrealistic.

Instead, it was recommended that development be undertaken as a series of incremental improvements. Based on this project, it was possible to plan and prioritise recommendations, allowing the activities that will have the most impact on the website (from a users’ point of view) to be completed sooner rather than later.

To that end, a roadmap was developed for the forthcoming website redesign, containing three distinct phases:

• Immediate actionsIn order to address website issues as soon as possible, an action list has been formulated. These are typically small changes, but will either address critical issues or set the scene for later improvements.
• For next redesignThe bulk of improvements to the website will require substantial changes to its structure and interface. These actions must wait until the next major site redesign.
• Long-term actionsSome areas of the library website are clearly tied to the fundamental way in which the library operates, or the attitudes of staff. Changes to these areas will take some time.

Conclusion

This project has shown that undertaking effective user research provides new insight into a website, and an institutional, audience. It is possible, and beneficial, to challenge an organisation’s established view of the world.

It is possible to challenge the organisation’s view of the world

The personas developed for the MUL website have since attracted the attention of staff in other areas of the Library. The benefits of defining clients in this way was apparent, and potential uses for the personas outside the domain of web design can now be seen.

However, the model used to describe the audience for a website can be different to that which departments within an organisation use. There may be valid reasons for doing this, and such an arrangement can work well as long as the purpose of each persona set is clearly defined and adhered to.

Having such a good understanding of the audience provides an excellent foundation for a subsequent website redesign. Information architecture, visual design and strategic decisions regarding content and functionality, can all make use of a ‘true picture’ of the target audience.

Finally, this project has shown the benefits of undertaking ‘needs analysis’ and a strategic review of a website. The findings from the review, beyond the audience model, gave tremendous insight into issues that surround the effectiveness of the MUL.

Acknowledgements

This project would not have been a success without collaboration between Macquarie University Library and Step Two Designs, in particular the efforts of Meredith Martinelli and the rest of the Library Learning and Development team.

Further reading

• An introduction to personas…
• In-house recruitment of users for research
• Start user research by talking with staff