"Although I was not initially ready to change over to the new business model proposed by Steve, i'm excited about working with Steve again to get my business going."
This real-life case study was the result of my final year at University, which was a human-computer interaction project to meet BCs accreditation. I contacted various local businesses through my connections and the business owner of West of Counselling services agreed for me to base my project around a website redesign.
This project met with standards; ISO 9241-210:2019 and ISO/TR 16982
I had to try and establish a good relationship with the client, actively listen and try and iron out as many details as possible while teasing out their unmentioned goals.Below are some of the main insights gained from the qualitative session.
To get an idea of where the problem lay with the clients existing product, I was granted access to historical data where I proceeded to extract the quantitative data. Along with some findings, an insight presented itself which was not expected. The clients business partner seemed to be getting more business, even though they were almost identical in terms of qualifications, therapies offered, user journeys and client positioning on the web page. I was presented with a lot of design questions and needed to formulate some hypothesis statements to create actionable steps.
We ran surveys (not statistically significant due to numbers in comparison to annual visitors) and based on the results, we know the hypothesis was valid if we got more than 70% of participants in a survey to choose the clients business partner on the existing website when given an identical task scenario to follow. We followed this up by asking the participants who chose the client over their partner, why they thought they did this. The attitudinal data indicated a positioning aspect (above partner) and there was more information. It was important to identify that due to psychological behaviour, users could have simply preferred a female counsellor. I had to check and recheck to avoid confirmation biases here by creating more hypothesis.
Mapping exercises were completed with the client to align all our mental models.
From a HCI perspective, I wanted to break each task down to inform the prototypes. The data was sourced from user interviews and a client-role play exercise in the absence of users (time sensitive at this point). I wanted to find out the red routes to incorporate into the wireframes.
After the client and I broke each user task down, it became apparent that the main business goal should be defined. A way of matching the needs of the target audience with the product - a way of linking it all together to meet business/user needs. One way of structuring a VPS came by way of a user-centred design canvas (Zawadski, 2019).
After several iterations, the client did not feel comfortable with the marketing aspect of a VPS and it was decided not to implement a message on the hero page at this stage. I advised the client that we could do a comparison test with a VPS and without with real users at the design stage.
Up to this point, we now required a structure to all the information (through experience, a content first strategy would be save time and effort later on) and this came about by conducting online open card-sorting exercises with users, using optimal workshop. Those results were triangulated with the data from a face-to-face hybrid card-sort with the client. These were not complicated due to the size of the organisation. I needed to link the relationships between pages/sections, where I commenced building a network of DoGo maps (click on image below).
At this stage of the project, I started to sketch out ideas on paper, which were then viewed on family and friends, however responses had to be managed carefully for any social desirability biases due to the connection between facilitator and participants. i decided to bunch any common themes together and get sign off from the client before commencing.
Using webflow, the prototypes were created to get buy in from the client. at this stage, it was obvious the client wanted something soon and to closely resemble a finished website. At this point, the client increased their feedback and revision changes in a structured spreadsheet.
After the 1st round of usability testing on the website was conducted, all results were categorised by severity ready for the iterative changes ahead of round 2 testing. 5 face-to-face semi-structured sessions were conducted using a think-aloud methodology.
at this stage of my learning, I used the Microsoft inclusive design toolkit to implement changes. No testing was carried out on real users regarding the accessibility of the product, however a designer evaluation - although restricted - ensured WCAG standards were met as a trade-off.
Communication: I would ask a lot more questions, using various techniques to elicit further information. On this project, I learnt very quickly that you get limited time with stakeholders and to use that time as productively as possible. Being cautious about asking lots of questions is a natural process I will go through.
Bias: I tried to suppress as much of my personal bias as a designer as possible on this project, however being inexperienced at this stage, I would introduce measures in future to prevent blind spots that restrict advantageous problem identifications.Checklists and awareness being the 2 important areas here.
Project management: In my opinion, too much time was taken up trying to perfect processes and deliverables. In hindsight, the process is messy and often requires enough data to inform iterations, where time can be wasted trying to mould a design only for it to be reshaped shortly after.
Goals: Being somewhat overwhelmed at times during the project, learning new skills and processes, it is vital to have clear goals and clear expectations throughout. It is through practice that my goal setting has improved by simply prioritising and effectively representing - be that usability, ux metrics, research or business, if you can set these out early, it prevents ambiguous dialogue and misunderstandings.
Iteration: I tried to build the perfect product with only a few years academic experience, perhaps lacking in confidence when it came to ux processes. Since then, I have heavily invested in testing as early and as often as i can to get the foundations correct. This was the most valuable lessons I continue to learn.
Client expectations: Working with evidence-led data, can conflict with client expectations. An example of this was producing a solution that addressed business needs - the introduction of a single therapist model - however unexpected challenges around the disruption of existing business relationships were not planned for, being the main reason behind an unsuccessful launch.
