A mindful audio layer for Google Maps that lets people walk a city by ear — staying present, rebuilding their sense of direction, and finally looking up from the blue dot.
Constant turn-by-turn guidance weakens spatial memory and keeps eyes locked to a screen.
Maps is visual-first. Its audio is purely functional — no layer that builds awareness of place.
Soundscape Mode: four audio profiles guiding by landmark, ambience, and haptics.
Reduced screen-checking in testing; reframed navigation from optimizing to exploring.
Maps have guided human movement for millennia — people read landscapes and built their own internal compass. GPS does that work for us now, and the convenience has a cost. Research links constant turn-by-turn guidance to weakened spatial memory and wayfinding skills (Fabrikant, 2024). Visual-first maps flatten a city into a line on a screen, leaving no room for the textures, rhythms, and sounds that help people orient, feel present, and connect to where they are.
U.S. Google Maps downloads in 2023 — the default way millions navigate every day.
Times per month the average person opens Google Maps. The behavior is constant.
Of surveyed users rely on GPS even for places they already know — the dependency is total.
A competitive teardown of Maps, Waze, and Apple Maps plus a user survey surfaced one consistent friction: people don't trust the screen, yet can't look away from it. Three findings shaped everything that followed.
Users need time to calibrate. The blue dot feels misleading when they're trying to figure out which way to face.
When phone-free, people orient by buildings, parks, and storefronts — not street names or compass directions.
Without signal, users fall back to paper maps or screenshots. The app abandons them when they need it most.
"I just want to confidently say I know where I'm going, without looking down at my phone every other minute."Maya, 24 — Everyday Urban Navigator · Primary Persona
Existing voice nav is purely functional — "turn left in 400 feet." It never helps you understand where you are. Synthesizing the research, the real problem wasn't accuracy. It was presence. That gap framed the brief:
A subtle audio layer for walking navigation. With the phone in your pocket, you follow the route by ear and stay present in the neighborhood. One size never fit how differently people navigate — so it ships as four switchable modes, each built on a research finding.
Low-key audio with only essential turn and transition cues. Keeps you oriented without overload — a subtle assistant, not a constant narrator.
Turn-based guidance anchored to real places — "face the park," "look for the museum entrance." Built directly on the 50% who navigate by landmark.
Ambient sound and short narratives about each place and the life around it. Reframes a walk from optimization into exploration.
High-contrast audio, visuals, and haptics for clearer orientation — stronger cues, haptic-first navigation, hearing-aid-friendly output.
The team split 50/50 on whether audio was even wanted. The insight: it wasn't audio that was the problem — it was the kind of audio. That pivot pointed straight at soundscapes.
Charted where Soundscape launches inside existing walking nav, so it felt like a native Maps capability rather than a bolt-on feature.
Prototyped the core surfaces — landmark cards, settings, the upcoming-landmark list, and a mini audio player — to pressure-test the model before polish.
Built all four modes in Google's visual language and ran task-based testing on real sidewalks to see whether people could tell modes apart and actually look up.
As a concept project, impact splits into what testing proved and what the feature is positioned to deliver at scale.
Landmark Mode measurably reduced how often participants looked at their screens during navigation tasks.
Users confidently differentiated every mode during task observation and found the headphone entry point intuitive to locate.
Immersive Mode reframed navigation as exploring a place rather than racing through it — a new emotional value for Maps' most-used surface.
A dedicated Accessible Mode opens confident walking nav to low-vision and hard-of-hearing users underserved by visual-first maps.
Testing flagged cue pacing and onboarding previews as the next refinement. Audio UX lives or dies on rhythm, not just content.
Accessible Mode needs to be built alongside the users who'll rely on it — co-design, not assumption. That's the priority for the next round.
The real bet is spatial memory recovering over time. I'd instrument KPIs and longitudinal surveys to test whether wayfinding confidence grows.