Canadian Association of Law Libraries 2013 Conference – It’s All About Redesigning to Stay Relevant

One of the big themes running through many of the workshops at this week’s CALL conference in Montreal was the redesign of products, platforms and processes. The conference ended yesterday.

The Monday session entitled “Please Don’t Make Me Think: User Testing a Faceted Search Engine” was about how the Centre d’accès à l’information juridique (CAIJ), Quebec’s Courthouse Library Network, conducts user testing sessions to validate the ergonomic and design aspects of many of its tools, including its new faceted search engine JuriBistro UNIK.

I served as a guinea pig at the session. I volunteered to go up on stage and “test” a music store website. I was told to try to order the latest Céline Dion CD for a Mother’s Day gift and failed miserably. The presenter then explained that she had deliberately chosen a badly designed website. My embarrassment (I was in front of a full room pretending to be interested in Céline Dion) served as a perfect introduction to the importance of usability testing.

During user tests relating to JuriBistro UNIIK, CAIJ made some very surprising discoveries about how people think about their searching, what they see and do not see on your site, and how they (mis)interpret symbols or language that appear perfectly clear to project members.

Lo and behold, they discovered that lawyers – their users – like to READ text and do not see or even understand many of the pretty and cute pictograms designers provided as shortcuts to very important features such as the fulltext of cases. So they replaced icons with text like “Texte intégral – Jurisprudence” and all the guinea pigs (sorry, I mean test lawyers) were happy.

Other features that test subjects overlooked or failed to see or understand included such things as sort options, the concept of “keywords in context”, links to help and search tips and the icon for further filtering of results which people thought looked like a martini glass (it was the Excel spreadsheet symbol for filter). Most of these items were replaced with text links.

The speakers, CAIJ’s Monique Stam and Anastasia Simitsis, User Experience Director with W.illi.am Digital Intelligence, used the Céline Dion example to emphasize a point.

User focus groups and surveys are insufficient to test (re)design ideas. Watching what potential users actually do when replicating real-life scenarios (e.g. trying to buy a CD for your mom) is required. The experience of listening to test participants think out loud as they try (and very frequently fail) to do stuff on the test site for your product or Intranet or online service is usually an eye opener (or a kick in the head) for project team members, designers and IT people.

They explained some of the possible flavours and techniques of usability testing, but they all offer similar benefits:

• identifying and fixing problems and ambiguous features early (“why does CAIJ have a martini glass on its results page?”)
• validating assumptions
• solving opinion battles among team members (web designer: “I think the filter pictogram should go here”; project manager: “No, I am sure the users will be more at ease if the filter is over there”; your boss: “I say the filter belongs next to the results”; of course, the test users are all going: “What’s with the CAIJ and all their little martini glasses on my results page?”)
• establishing a baseline to measure improvements
• reducing risk and reducing costs (it is way cheaper to fix problems at the earlier stages of (re)design than after launch).

Another Monday session on “Technology Project Management: Complexities and Challenges” offered three perspectives on how to manage large projects, and in particular the stakeholder expectations that are involved.

Julie Allard from the Quebec public corporation SOQUIJ shared her experiences with the design of the new product that will replace its AZIMUT Juris.doc legal information search tool in the fall of 2013. The project is highly complex, has incredibly short deadlines, involves a multidisciplinary team and clients are demanding. [SOQUIJ held a breakfast demo of its new search product on Tuesday morning. The coming changes can be easily summarized: extreme simplification of the interface. You search and find and manipulate results on just 2 screens. Everything happens on the first search screen and then on one single results screen. Full stop. Quite impressive.]

Getting the preliminary project planning right is key. This involves consulting customers and deciding what is and what is definitely NOT part of the project to maintain focus and avoid mission or project creep, probably the kiss of death of many projects. How many times does this simple truth get overlooked? Figure out what is NOT supposed to be part of the game plan and you will be happier.

Among the successful strategies Allard mentioned for not losing control:

• chunking the project into shorter cycles with precise interim deliverables to ensure the project team achieves small but constant victories along the way
• managing change requests – analyzing and documenting what their consequences would be and why they can be or cannot be implemented
• user testing, and quality assurance/control during the entire process to make sure any changes are in line with the needs expressed by customers (testing and quality assurance are too often overlooked until the very end)
• tracking problems and anomalies as well as tracking measures taken to address them

Managing stakeholders and their expectations requires a lot more thought than we often assume. In the case of SOQUIJ, they came up with a series of clear roles: a dedicated project manager, a project sponsor, a person to act as the representative of the customers, the project team, testers, etc. All clearly delineated roles.

Frédérique Tessier, project coordinator at Éducaloi, an award-winning website for public legal education, took the audience through the recent redesign of the site which included the complete reorganization of its content as well as migration to a new open source platform. The organization asked an outside firm to evaluate a list of prospective agencies who could manage the project in cooperation with Éducaloi members and external contributors (lawyer-writers). One was eventually chosen to lead the project.

Finding content seemed to be one problem with the old site. 10 users were recruited for a card sorting exercise to categorize the site’s hundreds of articles into a clearer classification scheme based on 10 broad topics (e.g. family), then broken down into nodes (“divorce”) and then into narrower areas (“child support”). On the page of any given article on the new site, there are links to other material within the same node and topic as well as in related nodes under different topics.

The final presenter at that session was François Montreuil, director of IT at Bibliothèque et Archives nationales du Québec (BAnQ). He focussed on streamlining stakeholder management during major technology projects.

BAnQ used to run projects with a 15-19 member project management committee to represent each sector of the institution, plus legal and the government. Meetings were long, unwieldy and many people were only there to defend their idea or turf. A new executive committee was created with 7 people that meets before the larger committee. It invites stakeholders to address it according to need and expertise and they are asked to leave after they have made their presentation. Actual projects are run by project teams with specific mandates and deadlines.

All three sessions presenters agreed there is no perfect structure for handling major technological projects. Whatever the structure adopted, clear communications with users and clear documentation and reporting are key elements.