Evaluation of the Hope Alliance Website
Hope Alliance is a non-profit based in Round Rock, Texas which assists survivors of domestic and sexual violence by providing shelter, services, and support. In the fall of 2016, I volunteered to evaluate this organization's website and provide guidance for its redesign, as they were struggling to reach their various target users and had a high bounce rate. Towards this end, I conducted a heuristic evaluation and user testing of the website, for which I recruited 5 participants. This project was the key turning point in my transition from a focus on libraries and archives to UX Research, as I started to view this field as a way I could help people.
I identified the viability of safety features, such as the pop-up on the home page and the global "Escape" button, as a key area for testing. During my heuristic evaluation, I discovered that the pop-up was verbose and served multiple purposes (note the donate button), which might be confusing to users.
Objectives
To determine whether the website's architecture, design, and functionality help or hinder various users of the site in accomplishing critical tasks:
- Potential Donors
- Finding a reason to donate
- Making both monetary and in-kind donations
- Potential Clients
- Finding information about available services
- Understanding and using current safety features (pop-up and escape button)
- Educators
- Finding information on issues for which the organization advocates
Methods
Heuristic Evaluation–I identified design errors using usability and accessibility heuristics from Nielson, Tognazzini, and Weinschenk/Sánchez. This evaluation informed the creation of my test script by helping me identify key tasks and provided guidance post-testing as I coded my test session transcripts.
Summative Testing–I tested for 5 critical tasks (identified in the Objectives section above) with 5 participants, which I recruited to represent a range of technical literacy. I recorded participants' time on task, number of errors, and level of completion and used the Think Aloud Protocol to capture participants' comments and thought process.
Participants were instructed to think aloud while completing each task. Having real users' opinions of the organization based on the organization and usability of its website helped me build an effective narrative and justify design changes to stakeholders.
Prior to the redesign, the "Donate Now!" button was hidden among underutilized social media links. No test participants used this button to make a monetary donation. The button has since been relocated to the top right corner of the home page, and the social media links have been moved to the top and bottom right corners.
Key Findings
Key information was getting buried for all user types–During my heuristic evaluation, I found significant issues related to feedback, navigation, unconscious processing, consistency/standards, and slow loading time. One particularly troubling problem, which also came up in all 5 testing sessions, was organizational in nature. A lack of cross-linking and inconsistent page naming made navigation cumbersome. Bulk text made pages difficult to skim. These issues led to negative feelings towards the organization and caused several participants to fail on 3 different tasks. In reaction, I discussed the need for progressive disclosure of information with stakeholders to prevent information overload.
Users had trouble making donations–The Call-To-Action button for monetary donations, the most important task identified by the organization's representative during our kickoff meeting, was difficult to find, so it took participants a minimum of 3 clicks to navigate to the donation form through an alternate route. Participants took a range of 2.5-4 minutes to complete the task, and one participant failed to complete the task altogether. Additionally, 3 out of 5 participants failed to find key information regarding in-kind donations.
One issue I found during my heuristic evaluation involved slow loading time. The webpages could take up to 10 seconds to load due to the way the underlying script was written. This lead to frustration during testing, as I had reload the home page in between tasks. I recommended that the scripts be amalgamated during the redesign, so that users don't get frustrated waiting for the website to load.
Challenges
- While I could've tested the website solely with fellow students, I set out to recruit a more diverse group of participants, as I wanted to test with people who had a range of technical literacy (low-high), educational backgrounds (some college-doctorate), annual household income ($26K-$150K), and age (23-62). My participants all fell into the "Potential Donors" group of users, as I was unable to recruit participants that fell into the "Potential Clients" or "Educators" groups. I still included the critical tasks for the latter two groups in my script.
- As this was my first experience running a usability test, I ran into some bumps during the first test, so I had to adjust my protocol for later sessions.