Audience Intelligence

See who's there. See who's watching.

Counts, dwell time and attention for every screen and zone — tuned to your space so the numbers reflect what really happened. Every measurement comes with a confidence score.

Four calibrated metrics, transparent definitions.

Every metric has a precise, documented definition. We publish what we measure, how we measure it, and the known limitations.

OTS (Opportunity To See)

Calibrated count of people with an opportunity to see the screen, within the display exposure zone, during the measurement window. Anchored to screen geometry and camera pose calibration.

Dwell Time

Duration an individual remains within the measurement zone, measured from first detection to exit. Reported as averages and bucketed distributions (0-5s, 5-15s, 15-30s, 30-60s, 60+s).

Attention

Proportion of dwell time where the individual is oriented toward the display surface. Reported as attention rate (0-1) and total attention events with confidence scoring.

Audience Window

A time-bounded observation period with aggregated audience metrics, confidence scores, and quality flags. Each window carries model_version, calibration_version, and event lineage.

Sample Dwell Distribution

0-5s
35%
Passers-by
5-15s
28%
Glancers
15-30s
20%
Engaged
30-60s
12%
Interested
60s+
5%
Captivated

Calibration requirements and standards alignment.

Accuracy depends on proper calibration. We are transparent about requirements and known limitations.

Calibration Requirements
  • Camera must be positioned with clear line-of-sight to the measurement zone
  • Screen geometry (dimensions, height, angle) must be configured during setup
  • Camera pose (position, angle, field of view) must be calibrated per deployment
  • Zone boundaries must be defined to delineate the display exposure area
  • Calibration should be re-verified after any physical changes to the installation
Known Limitations
  • OTS counts people in the zone, not guaranteed ad viewers — attention rate provides the viewing signal
  • Dwell accuracy depends on detection continuity; occlusion may cause under-counting
  • Attention detection uses head orientation, not eye tracking — confidence varies with distance and angle
  • Calibration confidence degrades if the physical installation changes without recalibration
  • Low-light and extreme weather conditions may reduce detection accuracy (reflected in quality_flags)
Calibrated OTS, dwell and attention measurement with full lineage on every record

100% edge processed.

Raw sensor feeds never leave the hardware device. Only calibrated, anonymous audience metrics are transmitted to the cloud.

Camera

Standard USB or IP camera feed

Edge Processing

Detection + calibration locally. Frames discarded instantly.

Calibrated Signal

OTS, dwell, attention with confidence and lineage.

Cloud API

Dashboard, API, forecasting, and partner integrations.

A fundamentally different approach.

Traditional analytics tracks demographics. We track calibrated audience presence and attention.

Feature
Demographic Tracking
SignIQ
Age & gender inference
Emotion recognition
Facial recognition
DPIA required
AI Act high-risk classification
Calibrated OTS counting
Dwell time analytics
Attention detection
Audience forecasting
Event lineage & model versioning
100% edge processing

Ready to measure
with confidence?

Calibrated audience measurement with transparent definitions, standards alignment, and full event lineage.