Point a camera. Measure the event. Tell the story.
Campaign Vision runs computer vision on your camera feeds to produce real-time crowd analytics — then writes the report for you. Here's the whole flow.
Point a camera
Connect a venue IP camera over RTSP, plug in a webcam, or stream from a GoPro. Most setups are working within minutes — there's nothing to ship or mount.
See live analytics
During the event your dashboard updates live: headcounts, dwell time, zone occupancy, density heatmaps, and capacity alerts. Share a view-only link with the client so they can watch alongside you.
Get an AI report
When the doors close, generate a narrative report written in the client's language — impressions, reach, engagement, peak times, and zone-by-zone performance — and ask follow-up questions in plain English.
Dig deeper: supported cameras · how we measure
Accurate counts, no surveillance.
Built for non-technical teams, honest by design — here's plainly what the system does and doesn't do.
It detects and counts people
Computer vision spots each person in the camera view and counts them — the real number on the floor, not a count of phones or an estimate from a clicker.
It follows movement, anonymously
Within and across cameras, it links the same visitor by body shape and clothing — not by their face — to measure dwell and map zone-to-zone journeys. Cross-camera matching is a capability for richer insight, not a guaranteed per-person track.
It keeps only the numbers
Frames are processed and discarded. What's stored are derived metrics — counts, dwell, zones, heatmaps — never footage of your visitors and no biometric identity.
It writes the report
AI turns the event's data into a clear narrative in the buyer's language, and answers follow-up questions over the data via chat — so the recap is ready when the event ends.
The full capability set.
Every metric the platform measures — core metrics every brief asks for, plus advanced capabilities framed exactly as what they are.
True headcounts
Count actual people in frame — not phones or devices. The reliable visitor number every brief asks for first.
Dwell time
How long visitors spend at a booth, display, or zone — so you can show engagement, not just footfall.
Zone engagement
Draw zones on your space and get per-zone visits, dwell, and peak times. See which corners of the activation actually work.
Impressions & reach
Walk-by impressions and engaged visitors, reported in the language brands buy in — reach, frequency, and engagement rate.
Density heatmaps
A live and post-event heatmap of where crowds form, so you can place staff, product, and signage where the people are.
Capacity & queue alerts
Set a capacity per zone and get real-time alerts as it fills up — useful for safety, staffing, and queue management.
Cross-camera journeys
Anonymously link the same visitor across multiple camera views to map how crowds move between zones. A capability for richer journey insight — not a guaranteed per-person track.
Attention signal
An estimate of whether people are facing the brand or display — a directional read on visual attention, not just presence.
Estimated demographics
Optional aggregate age-band and gender mix for the crowd — never per-person, no one identified, off by default and enabled per event with explicit sign-off.
AI reports & chat Q&A
Auto-generated, client-ready narrative reports — plus a chat you can ask questions of in plain English ("What was peak dwell at the bar?").
Want to see it on a setup like yours?
Book a demo and we'll run through the live dashboard and a sample report for an event close to what you produce.