Designing hearing support without medicalising it

If you build an app that touches hearing, you immediately walk into a regulatory minefield. The medical-device label is fuzzy on the edges, and the safest position is to assume anything you say about “improving” or “treating” hearing puts you in a regulator’s notebook. That’s the right default — but it’s also a design constraint, not just a legal one.

HearLab is explicitly a non-medical companion app. That positioning is real, not just a disclaimer. It shapes what we build and what we don’t.

The framing: “what just happened?”

Hearing aids, cochlear implants, and assistive listening setups already exist. The clinical layer is staffed. What’s missing for many users is a way to log and remember what just happened in a given environment — the restaurant that was unmanageable, the meeting that worked surprisingly well, the headphones that sounded weirdly thin on the train.

A non-medical companion app can do exactly that without claiming any diagnostic role. The job is:

• Capture context (location type, signal sources, sound levels, fatigue).
• Make it cheap to log (two taps, optional voice note).
• Surface patterns weekly.
• Hand the user a summary they can show their audiologist.

That is genuinely useful, and it does not touch any clinical decision.

The needles to thread

1. Regulatory

Avoid anything that implies treatment, diagnosis, fitting, or correction. Even language like “calibrate your hearing aid” is risky. Stick to logging, awareness, captions, and routing visibility. When in doubt, frame it as the user’s experience rather than the device’s behaviour.

2. Social

A “hearing app” that screams disability fails. Many people who would benefit don’t identify with the category yet, or don’t want to broadcast it. So the affordances need to be quiet — the check-in shouldn’t feel medical, the captions shouldn’t feel clinical, and the dashboard shouldn’t look like an EHR.

3. Technical

On Android, the routing layer (AAudio, MediaRouter, BluetoothAdapter) hides a lot of what users want to see. Surfacing “your audio is currently routed to your hearing aid via LE Audio” is more useful than people realise — and it’s not exposed cleanly in any consumer-facing app today.

4. Business

The hardest one. End users don’t reliably pay for accessibility apps. The pattern that works is a B2B2C layer: a free-or-cheap companion app for users, a dashboard layer for audiologists and hearing centres who want anonymised insights from their patient base.

What we don’t do

• We don’t fit, adjust, or recommend hearing aids.
• We don’t score hearing.
• We don’t market to clinical buyers using clinical language.
• We don’t treat captions as a “feature” — they’re a baseline.

What we do

• We help people notice their hearing experience.
• We help them describe it.
• We make it cheap to share with the professional in their life who has the clinical context.
• We treat hearing accessibility as a serious, design-led product surface, not a CSR project.

If that framing matches a problem you’re working on, we’re looking for design partners.