Measure participating-device exposure beyond the delivery log.
Phantone records qualifying acoustic detections from participating apps when signaled television or streaming media plays nearby—providing a direct measurement layer alongside panels, server logs, and existing attribution methods.
Here's what a broadcast campaign looks like with Phantone.
- T+00:00 · STEP 1 / 5
The advertiser or network applies the signal to the TV spot or ad break before it goes to air. The audio is unchanged. The signal is in.
- T+00:12 · STEP 2 / 5
The spot airs nationally — on cable, satellite, OTA, CTV, or streaming. Participating apps in a supported listening state can detect the signal through the device microphone.
- T+00:30 · STEP 3 / 5
The SDK decodes the signal on device. The impression is logged — campaign ID, creative ID, timestamp, and a privacy-preserving device or installation identifier — and delivered to the advertiser's endpoint.
- T+01:30 · STEP 4 / 5
The advertiser's dashboard shows a feed of qualifying detection events — each one timestamped, campaign-attributed, and backed by a W3C Verifiable Credential that sits alongside your existing measurement.
- T+10:00 · STEP 5 / 5
A second-screen experience fires on participating devices: a product offer, a 'learn more' unlock, or loyalty credit — timed to the moment the spot ends.
What the viewer experiences.
Appears after the spot, inside the participating brand or partner app.
What you see.
The problems this vertical wants solved.
- 01
A direct count of participating devices that detected a specific signaled spot.
- 02
Signed, verifiable impression records their agency can audit independently, without relying on our database.
- 03
A second-screen unlock timed to the end of the spot, without any user interaction.
Framed in Phantone's own voice from partner conversations — not attributed quotes.
Add direct acoustic evidence to broadcast attribution.
We work with networks, advertisers, and agencies to deploy on existing broadcast infrastructure — as a measurement layer alongside the panels, server logs, and attribution models you already use.