Gold Coast Health

The Process

Desktop research

How can we collect measurable data to learn user behaviour?

I began this journey by analysing the metrics in Google Analytics. I discovered an interesting behaviour,

• 70% of users don't click past the homepage

• Popular pages totalled an estimate of 6% majority views. Why was this percentage of majority views so low?

• The top accessed pages were services, appointment information and contact numbers.

Is there past research we can use?

We know Queensland Health went through a major CMS migration to offer a 'Out of the Box' service for healthcare organisations that don't have a dedicated Digital Team to design templates.

• I contacted the head of content strategy, who shared their extensive research and personas; this helped us learn the best type of users to target for our research.

Who are the healthcare leaders, and what trends can we leverage for our site?

I researched the top hospitals worldwide and in Australia. Although the best hospitals didn't always display great website designs, we identified trends in navigation, taxonomy, and features across the board.

Stakeholder interviews

Before we could talk to our users we needed to identify the research goals.

• What do decision makers and SME's expect from this redesign?

• Are there any major business frustrations that the website can appease?

• What does the current site not do well?

We interviewed 42 stakeholders across the board, all from different departments.

A total of 2100 hours of interviews.

Date synthesis & analysis

Using the process of data clustering, I grouped similar topics under a label identifier, which helped me refine the data to stakeholder goals, needs, pain points, and jobs to be done. I analysed the data to draw insights to help guide the research goal.

Findings

I identified the top insights from stakeholders, which helped guide the questioning and goals for the user interviews.

• Stakeholders felt the website design didn't portray Gold Coast Health's brand values of 'Always care. '

• There was not enough information about the emergency department process and services, resulting in high traffic to emergency services.

• Content is unreliable and out of date.

• There is a large amount of calls to Switchboard; users have questions about their appointments.

User Interviews and Usability testing

We need to learn who are the main users of the website and what are their main tasks. Is there an opportunity to redirect traffic to the website to help users self serve?

We identified the key personas to talk to:

• Patients experiencing an emergency

• Patients in Outpatients

• Patients in Inpatients

• Patient carers

• Jobseekers

User surveys

We were fortunate to partner with Griffith University which volunteered to recruit users to participate in a survey. I defined the questions to collect measurable data based on the quantity of preferences belonging to which demographic.

Findings

• The majority of users use their mobile device to access the website.

• The majority of users between the ages 25 and 40 have accessibilities needs (colour blindness and text size)

• 74% of users accessed the website to locate more information about their appointment.

User interviews

With all the information we have collected up to now, we were confident we had a clear goal for our research.

We spoke to 5 users per persona, totalling 25 interviews.

A total of 750 hours of interviews.

Data synthesis & analysis

Using the thematic method approach, I coded each snippet of raw data, grouping and clustering accordingly to refine the groupings. After thorough analysis, we reached a list of key insights.

Findings

We refined the list of insights, keeping in mind all the data we have gathered at this point.

• Patients and carers don't feel supported throughout the patient journey, specifically post-care.

• Patients utilise the website to manage and seek more information about their appointments.

• Patients look to the website to find contact numbers for a service.

• Jobseekers expect to see a list of job vacancies on the website.

• The content is text-heavy, which proves to be hard to scan and digest.