- Mobile UI/UX
- Accessibility
Artour
A self-initiated mobile concept for gallery audio tours, designed around accessibility
Museum apps offer audio tours in one language and call it accessibility. I designed for the visitors that leaves out.
Role
Solo: research, UX, UI, testing
Timeline
July – Aug 2023
Type
Self-initiated concept
Testing
Two rounds, 5 participants
0 of 3
Audited apps offered audio translation
2
Rounds of usability testing
5
Participants, ages 19–60
Jump to a section
Overview
A self-initiated project asking what a gallery audio tour would look like if accessibility were the starting point rather than a settings menu.
I ran this end to end on my own: competitive analysis, interviews, both rounds of usability testing, and all of the design. Nobody briefed it, which meant the framing was mine to get right or wrong.
The premise: galleries increasingly ship apps with audio tours, and those tours are overwhelmingly monolingual. For a visitor who doesn't speak the language fluently, the app technically has an audio tour and practically doesn't.
Three gaps, one pattern
Accessibility treated as an add-on
Language support and captions appeared inconsistently across the apps I looked at, and audio translation appeared nowhere at all.
Virtual tours misread as a substitute
Galleries build virtual tours as a replacement for visiting. Visitors were using them as a way to decide whether to visit, a different job with different requirements.
Nothing works across galleries
Every institution ships its own app, so the learning cost resets at every door.
I audited three apps (the Walt Disney Family Museum, the Art Institute of Chicago, and the St. Louis Public Library), looking specifically at language support, closed captions and search. Only a few offered multi-language support or closed captions. None offered audio translation.
Five people, deliberately mixed
Participants aged 19 to 60, recruited so that the accessibility needs weren't hypothetical.
- Non-native English speakers
- 3 of 5Non-native English speakers
- Regular accessibility feature users
- 3 of 5Regular accessibility feature users
- Frequent gallery visitors
- 2 of 5Frequent gallery visitors
“Non-native English speakers described feeling locked out of the full experience even when an audio tour was available.”
The second finding reframed the virtual tour entirely: frequent visitors weren't using virtual tours as a substitute for going. They were using them as a decision-making tool, checking whether a show was worth the trip.
That changes what the feature needs to do. A substitute optimises for completeness; a decision tool optimises for a fast, honest impression.
Accessibility wasn't the edge case
“The users who needed accessibility features most weren't a niche edge case. Designing for them produced a better experience for everyone in the study.”
Serial codes made this concrete. Most gallery apps ask you to find a number on a placard, type it in, and check it matches. That's a small annoyance if you're fluent and sighted, and a genuine barrier if you're neither. Removing it improved the flow for every participant, not just the ones who struggled with it.
Two flows, three features
I mapped in-person and virtual visits as separate information architectures rather than forcing one through the other, because the two have genuinely different goals.
Multilingual audio with synced subtitles
Translated audio, captions that track it, and adjustable text size. This is the feature the whole project exists for.
Image scanning instead of serial codes
Point the camera at the artwork and the app identifies it. It removes the typing barrier and, because it doesn't depend on a gallery's numbering scheme, it's the piece that could work across institutions.
Chaptered information architecture
Content organised so visitors set their own depth: a casual visitor gets a short read, an engaged one keeps going, and neither is forced through the other's experience.
What testing changed
Low-fidelity testing confirmed the flows held. High-fidelity testing found two real problems.
Only 2 of 5 could find the language toggle
The single most important control in the app, and most people couldn't locate it. I centralised the key controls into one action bar. Discoverability failed because the controls were scattered, not because they were badly labelled.
2 of 5 asked for filtering
Participants were scrolling through unfiltered content. I added a hashtag-style category system that works across both Explore and Liked Works.
The honest gap
Working solo sharpened one thing in particular: when you're both the researcher and the designer, you develop a much faster instinct for which insights are actionable and which are merely interesting.
Next steps would be dedicated research with non-native English speakers, validation of the iterated action bar and filtering, and an exploration of what AR adds to the virtual tour, if anything.