retail analytics

Digital signage analytics that measure the audience, not the individual

In short

Digital signage audience analytics measures how many people passed a placement, how many were positioned to see it and how long they dwelled nearby. Composition can be reported in aggregate using Ayonix age and gender estimation, or omitted entirely in anonymous mode where volume and dwell answer the question.

Request a pilot Send a sample video

Conceptual illustration of checkout queueing used for wait measurement. AI-generated conceptual image, not a customer deployment.

Scope

What this analytic does, and what it does not

What it detects

Passing traffic volume at a placement
Opportunity-to-see count for people within the viewing zone
Dwell time within the viewing zone
Aggregate audience composition, where estimation is enabled

What it does not guarantee

That anyone actually looked at the screen — proximity is not attention
Any measure of recall, persuasion or purchase influence
Individual-level audience records — the output is aggregate only
Accurate composition where the camera cannot see faces near-frontally

The problem

What customers are actually dealing with

These are the situations that lead teams to look at this analytic in the first place.

Signage inventory is sold on a footfall estimate that was never measured at the placement itself.
Advertisers ask for audience evidence and the operator has only building-level totals.
Screens are placed where there is traffic but no reason to pause, and nothing measures the difference.
Demographic reporting is requested without anyone establishing a lawful basis or considering an anonymous alternative.

How it works

Detection workflow

Every event carries the rule that produced it, so an operator can see why they were alerted.

1

Define the zone

A viewing zone is drawn in front of each placement, sized to the screen and its realistic viewing distance.
2

Count and track

Passing traffic and people entering the viewing zone are counted and tracked.
3

Measure dwell

Time spent inside the viewing zone is accumulated per anonymous track.
4

Estimate composition

Where enabled, aggregate composition is estimated with an uncertain category and minimum-count suppression.
5

Publish

Per-slot aggregate metrics are published for campaign reporting; no per-person record is retained.

Digital signage audience analytics workflow diagram

Configuration options

Viewing zone geometry per placement, matched to screen size and viewing distance
Minimum dwell before a person counts as an opportunity-to-see
Reporting slot length, commonly aligned to campaign scheduling
Whether composition estimation is enabled or the placement runs in anonymous mode
Minimum observations per slot before any figure is published
Baseline periods for campaign comparison

Alert quality

How irrelevant and duplicate alerts are reduced

No video analytic eliminates false alerts. These are the mechanisms that reduce them, and the residual rate is measured on your own cameras.

A minimum dwell threshold separates people who passed from people who were plausibly exposed.
Slot-level aggregation with minimum-count suppression prevents thin data being reported as a finding.
Anonymous mode is the default where composition is not strictly required.
Consistent zone sizing across placements is what makes inventory comparable.
Proximity metrics are a measure of opportunity, not attention, and are labelled as such in reporting.

Prerequisites

Camera requirements and environmental limits

Camera suitability decides more of the outcome than any software setting. These are assessed per camera before commitment.

Camera requirements

A view covering the viewing zone in front of the placement, not the screen itself
For composition estimation, a near-frontal view of faces at adequate resolution within the zone
Even lighting in the zone, which is often difficult directly in front of a bright screen
Stable mounting so zone geometry stays consistent across a campaign
Sufficient resolution at the far edge of the viewing zone

Environmental limitations

Screen glare and the brightness of the display itself can make the area in front of a screen a difficult exposure environment.
Proximity does not establish that anyone looked at the screen.
Composition estimation is unreliable where the camera view is oblique or overhead.
Placements in transit corridors will show high volume and low dwell by their nature, which is a property of the location rather than a fault.

Delivery

Alerts, evidence and where they land

Event and evidence fields

Placement and viewing zone identifier
Slot start and end
Passing volume and opportunity-to-see count
Median dwell in zone
Composition where enabled, with uncertain proportion
Suppression flag

VMS integration

Signage analytics is a commercial reporting output rather than a security event and does not enter the VMS alarm queue. It is normally delivered to the signage or campaign management platform.

See compatibility states →

Deployment options

Edge processing at the venue so only aggregate slot metrics leave the site
On-premise processing where placement cameras already aggregate locally
Anonymous mode as the default configuration where composition is not contractually required

Compare architectures →

Measurement

Dashboard metrics and pilot acceptance criteria

Acceptance thresholds are agreed with you before the pilot starts. This site publishes no benchmark figures, because they do not transfer between sites.

Dashboard metrics

Passing volume and opportunity-to-see per slot per placement
Median dwell in the viewing zone
Placement ranking by opportunity-to-see and by dwell
Campaign period versus baseline
Suppressed slot count
Camera availability per placement

Measurable pilot criteria

Passing volume accuracy against a manual count at sampled slots
Opportunity-to-see counts checked against observation at the placement
Dwell accuracy against stopwatch ground truth in the viewing zone
Exposure quality assessment in front of the lit screen at night
Confirmation that suppression behaves correctly at quiet times

Governance

Privacy and governance

All outputs are aggregate per slot. No individual audience record is created or retained.
Where composition estimation is enabled, every constraint on aggregate demographic analytics applies, including the exclusion of sensitive attributes.
Anonymous mode should be used wherever the commercial question can be answered without estimation.
Signage-adjacent cameras require the same signage and lawful-basis treatment as other surveillance cameras, and often more scrutiny because the purpose is commercial.

Frequently asked questions

Does this prove people looked at the screen?

No. It measures presence and dwell within a viewing zone, which is an opportunity-to-see measure. Gaze measurement is a different technique with different camera requirements, and it is not offered here. Reporting labels these metrics as opportunity rather than attention.

Can we report audience composition?

Only in aggregate, only with an explicit uncertain category, and only with minimum-count suppression. Every limit described on the aggregate demographic analytics page applies. Where the contract allows, anonymous mode is recommended instead.

Why is the area in front of a screen difficult for cameras?

A bright display in the frame drives automatic exposure down, which can silhouette the people in front of it. Placement and exposure settings are assessed during the readiness review, and some positions will need a different camera angle before they can support reliable measurement.

Keep reading

A controlled pilot establishes what this analytic actually does in your environment, against acceptance criteria we agree before it starts. Send a sample video first if you would rather see an assessment before committing to a pilot.