Skip to content
Get a demo

Hybrid mobile ACR

Audio fingerprinting + watermarking, engineered for the smartphone. One device detects real ad exposure across CTV, Linear, BVOD, YouTube, Audio, and OOH.

Mobile ACR has historically traded accuracy for scale. Beatgrid's hybrid architecture delivers both — passive, battery-efficient, and reliable in real-world listening environments. This is the detection engine behind the Human Exposure Graph.

Get a demo
waves_room4

How It Works

The world's first hybrid audio recognition: Fingerprinting + Watermarking

Audio Fingerpriting

Audio fingerprinting

Fingerprinting detects advertising by matching audio signatures against Beatgrid's reference database. It enables consistent ad detection across TV, CTV, streaming, YouTube, radio, and digital audio — without modifying the creative, and retrospectively against campaigns already in market.


Fingerprinting answers one core question: Which ad was the person exposed to?

Audio Watermarking

Audio watermarking

Beatgrid's proprietary watermark embeds a lightweight signal into the ad's own audio to identify the channel it played on. Because the signal travels inside the creative, no broadcaster integration and no platform permission is required — the ad is encoded once, and detected wherever it runs.


Audio watermarking answers one question: which channel was the ad viewed?

Award-winning Technology

Why this algorithm changes measurement

Mobile devices impose hard constraints: battery, background-processing limits, and environmental noise have historically capped the accuracy and scale of smartphone-based ACR.

Hybrid recognition removes those trade-offs by splitting detection across fingerprint and watermark signals. In practice that delivers:

  • Exposure precise enough to serve as ground truth
  • Continuous passive detection that panellists tolerate
  • Reliable detection in noisy, real-world environments

This is what makes large-scale passive cross-media measurement operationally viable.

Human data - animated

Passive tracking. Built to scale.

phone_animation

Low battery impact

Optimised audio sampling runs at under 0.5% battery. Passive detection continues across long panel sessions without panellists noticing.

Built for continuous scale

Detection runs within mobile background-processing limits, adapting to device and battery state so it stays active across the day. It persists without the app being open and without any action from the panellist.

Smartphone-based cross-media measurement

One person. Every screen, every channel. The smartphone becomes the universal meter for the entire media landscape.

The Panel Behind it

ACR detection is one half. The panel is the other half

Hybrid ACR produces detections. Run continuously across the same single-source panel — the same people, every screen, every day — those detections become a person-level record of real advertising exposure. That record is the Human Exposure Graph.

  • 220K active metering panel
  • 100% 1st party data
  • Over 5.8 M installs
Explore the ACR Panel →
Human data

Privacy-first by design

Beatgrid’s ACR technology is built for a privacy-conscious measurement ecosystem.

No recording of conversations

No recording of any audio or conversations

Signal

Audio is converted into audio signatures

data transfer

On-device signal processing minimizes data transfer

Explicit panelist consent and full opt-out control

Explicit panelist consent and full opt-out control

We identify advertising exposure — not personal behaviour or spoken content. GDPR and CCPA-compliant and an IAB Europe TCF-registered vendor.

Case studies

Beatgrid AI

Human Data for AI-Driven Advertising

Media planning and optimisation are increasingly run by algorithms, yet many AI systems still train on modelled data from external DSP or media platform-reported metrics, compounding modelled error. Beatgrid provides real, deterministic human-verified data to increase accuracy and prevent model drift and enables: 

  • Media allocation decisions
  • Cross-platform frequency calibration
  • Creative performance learning
  • Outcome-based campaign optimization
Learn About Beatgrid AI
Human data - animated

NEW: Beatgrid AI

Beatgrid's exposure data is also being used to calibrate AI media models — now in closed testing

Learn about Beatgrid AI

Frequently asked questions

What is Beatgrid's ACR technology?

Smartphone-based automatic content recognition that detects real ad exposure across CTV, Linear, BVOD, YouTube, Audio, and OOH — on the same person, building the Human Exposure Graph.

What is “hybrid mobile audio recognition”?

Beatgrid uses a hybrid of audio fingerprinting and proprietary audio watermarking. Fingerprinting answers “which ad was the person exposed to?” by matching audio signatures against our reference database. Watermarking embeds a lightweight signal into advertising audio to differentiate channels and improve detection certainty. Together they deliver higher exposure certainty with lower device resource consumption.

Why hybrid fingerprinting plus watermarking?

Mobile devices have unique constraints — battery limits, background processing rules, and noisy real-world audio environments. Fingerprinting alone struggles in those conditions. The hybrid approach delivers higher exposure certainty, lower device resource consumption, and greater robustness in real-world listening environments — making large-scale passive cross-media measurement operationally viable. 

Together, they create unparalleled exposure certainty while lowering device cost.

Does the app record what is happening around the user?

No. I capture a mathematical audio signature and discard the audio immediately. No conversation or content is ever recorded or stored. The user's phone data storage will confirm that. On-device signal processing minimises data transfer, and the system identifies advertising exposure.  Panelists provide explicit consent and have full opt-out control.

How is the SDK optimized for mobile?

Audio sampling is tuned to run at under 0.5% battery with 2–5 second detection, enabling continuous passive measurement across long panel sessions.

It's built for scale, to run across hundreds of thousands of opted-in panellists simultaneously while maintaining statistical integrity.

Why is one smartphone enough to measure every channel?

Because the phone travels with the person and because audio is the universal carrier across CTV, linear TV, YouTube, Social Video, streaming, radio, and digital audio, a single optimised device can detect exposure on every screen the panellist encounters — including out-of-home environments such as bars and hotel screens.

One-meter device. Every channel. The smartphone becomes the universal measurement device for the entire media mix.

How does Beatgrid's ACR support AI-driven advertising?

Media planning, bidding, and optimization are increasingly algorithmic — and AI systems are only as good as the signals they learn from.

Beatgrid provides verified, deterministic human exposure data for media allocation decisions, cross-platform frequency control, creative performance learning, and outcome-based campaign optimisation.

Trusted exposure data becomes the human-verified ground truth that calibrates AI planning and MMM, rather than training them on modelled estimates.