A retail client asked us:
What usability challenges and opportunities exist for our AI customer service chatbot among users with accessibility needs?
Our client reached out to us for help conducting an accessibility study on the latest version of their customer support chatbot. The team had already run similar usability studies on key tasks with larger swaths of their customers, but wanted to take the next step to see how their AI chatbot could work for those who use assistive technologies.
Up to 1 in 4 Americans have some form of disability, and in order to design inclusively and access a larger market share, they needed to ensure the chatbot was accessible in its current state and understand what opportunities there were for future iterations.
We Answered:
Since this was one of our client’s first accessibility studies, bringing all stakeholders along the research journey to increase empathy and understanding was hugely important. We crafted an end-to-end study that would test both the usability of the current live chatbot, as well as probe for potential improvements in future iterations.
We kicked off the project by recruiting:
- two key customer segments who shopped on their site
- a mix of participants with vision, motor, and cognitive impairments
After learning the ins and outs of the chatbot’s role and potential interactions, we conducted hour-long in depth interviews with over 30 participants, adapting along the way to challenges that arose from recruit difficulties to assistive technology hiccups.
Outcome:
To ensure all stakeholders and teams benefited from the research, we delivered a story-driven report detailing the challenges and opportunities uncovered during the sessions. Findings focused on:
- how their key segments with varying impairments and ranging usage of assistive technologies felt completing key tasks
- where they felt the chatbot could better meet their needs
To help the client better plan for future development, we surfaced insights around existing sentiment towards AI within the accessibility community. Some findings were surprising and unexpected, but offered clear actionable next steps for improvement in both the short-term and long-term.
...
Learn more about UX research for AI-powered experiences.
