UX redesign doesn’t need to have massive changes
Chapter 2: The story continues to get bigger, with more contributors joining the picture.
Week 2 of the course, signified the start of Project 2, to be completed in 2 weeks by 2 people. It would be the first project where we’d be working in teams, and teamwork added an extra layer of complexity which i was apprehensive of, in the event that I couldn’t get along with my partner.

Camilla was my designated partner by the sorting bag, a teammate and friend I came to know and appreciate over the course of this journey we completed together. To be honest, I was slightly hesitant to work with her at the start, as she was one of the few people I had barely spoke to in class, and had no idea what to expect. However, my fears were quickly dispelled as she proved to be a contributive partner who would pull her own weight in the project, ensuring that tasks were done even if it meant working late at home, and even was willing to help other groups with their Axure questions (which was her strong suite). I believe we balanced out well and adapted to each other’s working style, she being the caring and creative designer with a knack for keeping the atmosphere light, while I was the pragmatic researcher, who sometimes was a little ‘kan chiong’ (Hokkien word for nervous). Thanks Cam!
The Project —
The project brief dictated that we had to evaluate and redesign a designated college website — NUS Business School, with the UX design process we were being equipped with. The tools we were to use was taught during the two weeks while we carried out our project, hence it was a ‘learn and apply’ method, and was probably one of the contributing reasons to my feelings of being slightly ‘lost’ in this project’s process. As I was used to creating project plans knowing what was required in each step, the uncertainty of not having learnt all the skills and tools we needed for the project meant the work plan was incomplete (which gnawed away at my inner OCD self).
Since there was 3 pairs revamping the same school site, we decided to work together because the more brains there are, the faster it’ll be, right?

Now I understood why Agile process teams should be as small as possible. With six people working together on this project, we made slow progress on the first two days as there was competing thoughts and different perspectives on how to approach the project. Each of the individual teams had noticed this as well, and the three groups went our separate ways soon after completing the initial heuristic evaluation and content audit.
Research
Once it is seen, can’t be unseen.
As UX designers-in-training, we started to develop a lens over our eyes, where every website I look at now, I start to notice little details or elements relating a the user experience and interactions, and subconsciously analyse that website. In the UX world, the fancy term for it is conducting a Heuristic Evaluation (HE).
My mega group had decided to conduct the HE using the Jakob Nielsen’s Usability Heuristics framework, as a way to objectively review the website quantitatively, which also allows for the competitor websites to be compared on the same scale. Personally, I did not find the quantitative scoring method of HE to be that useful for the rest of the project, as the scores did no more than illustrate the ‘type’ of problem the website needed the biggest help revamping. The 10 heuristics would be more useful as ‘tags’ to identified problems, for organising the problems found into categories.

At first, I wanted to choose an overseas university for comparison (Harvard Business School, Wharton School at UPenn), but realised that their undergraduate business programmes were structured differently, hence it would not have had the same browsing pattern and hard to make a fair comparison. In the end, we went back looking at local universities and chose Singapore Management University, and Singapore Institute of Technology, for they both had similar degree programmes, and possibly the same pool of prospective students.
Post-HE, we realised there was an information gap — we needed to know how actual users used the website (do they think of going to the programmes section first? Or admissions requirements first?) Thankfully Cam had a few helpful minions cousins within our target persona’s age group, whom we sent out a simple Google form survey with four questions to give us some insights on their exploratory habits thinking about university.
We were slightly lost at the next step at this stage. Like what’s next? Card sorting it was, to understand how users relate and organise information in their heads. At this point we were just feeling our way through the process, thankful for our instructor Nuno’s advice/reminder that we should do the card sorting exercise ourselves first, and due to the limitations of the card sort, should not go further than three levels, which would be a too overwhelming exercise with too many cards.

Cam and myself sorted the cards physically with post-it notes at first, which we took more time than expected, trying to put ourselves into our users’ shoes. Still, by forming the groups with the card sort, we had also managed to create our draft site map, version 1.0.
The next step was taking the card sort online, to validate our assumption on whether users thought the same way we did. To a certain extent, they did, but the real gem of the card sort lay in the smaller differences, exposing some groups that needed restructuring.

Armed with the draft IA version 2.0, we tested the user paths in a different way — using the Tree-test, to check if users could achieve certain tasks that we set out for them to find. The results were surprisingly bad, with three tasks achieving more than 50% failure in finding the right ‘end goal’. That meant there must have been something not right during the tree test. Further digging with the analysis actually revealed that the problem lay in the phrasing of the task questions, some of the category phrases, which we then renamed in our revised IA.


Having done up the sitemap (and revised multiple times), we drew out the user flow for all three personas (demarcated with different arrow colours). Wireframe sketches was the fastest way for us to communicate our ideas to each other, and easily changeable. After having done sketches for two projects, I can’t deny the usefulness and helpfulness of it, and is definitely a step I will not skip at all.

The Prototype was done directly in Axure RP, as we decided to save on time by skipping the mid-fi version of the prototype. While we learnt the skills we needed in class before, creating a working prototype and troubleshooting was a lot more time-consuming.
One key takeaway I felt, was that we did not organize findings clearly during the research and testing stages, it was time-consuming to return to the notes to do calculations for the report/powerpoint.
List of changes we made to the NUS school biz site:
- IA reorganisation
- IA — shortened links + restructured
- Removed redundant pages (Undergrad page, Overview, shortened links)
- Undergrad label (header) = visibility
- Sticky header
- Side nav bar (coloured current page)
- Breadcrumbs
- external links symbol
- Main page layout (cleaner, simpler) — added carousel
- Prospective student page (layout — info visibility)
- Consolidated Login links for easy access on header
- Added filter widget to scholarships page
Overall learnings
- There will always be doubts, so best to clarify them if possible. if not, learn to live with them
- Presentation: need to strengthen the connection between before and after process. We cut out too much in order to fit the presentation within the time limit.
- Good teamwork is really important for a successful project (and makes life so much easier)