Retail analytics
Retail Video Analytics: CCTV Analytics and Computer Vision in Retail, Explained
Retail video analytics is the use of computer vision on CCTV footage to turn what a camera sees into counts and events: people entering, paths taken, time spent in a zone, queue length, an advisor approaching a customer, a hand reaching a shelf. CCTV analytics is the older name for the same thing; computer vision in retail is the technique underneath it. It is the branch of retail analytics that sees the most and hears nothing.
What video analytics can measure
- Entries and exits, usually more accurately than a break-beam counter, with staff excluded by zone or uniform.
- Paths and heat maps: which zones visitors pass through and in what order.
- Zone dwell: how long a visitor stayed in front of a category or display.
- Queue length and wait at billing or the service counter.
- Approach events: an advisor reaching a customer, and how long after entry.
- Shelf interaction: a product picked up, examined, put back.
- Loss prevention and safety events, which is where most CCTV analytics budgets began.
Where it sits among the other branches
Video analytics is less a branch of its own than a data source for several others. It feeds footfall analytics with a better count, customer journey analytics with paths and dwell, merchandising analytics with shelf interaction, and workforce analytics with approach times. Most Indian retailers already own the cameras; the analytics is a software layer on the existing recorder or a small edge device in the store.
The unattended visit is the most useful thing it adds. A camera can show that a customer stood at the counter for four minutes, was not approached, and left. No counter, POS or CRM records that visit at all.
Video metrics: definitions, sources and owners
| Metric | Definition | Source | Cadence | Owner |
|---|---|---|---|---|
| Entries | Visitors crossing the entrance line, staff excluded | Entrance camera, vision model | Daily | Store operations |
| Zone dwell | Median time per visitor in a defined zone | Floor cameras, vision model | Weekly | Merchandising, store operations |
| Path share | Share of visitors who reach each zone | Floor cameras, vision model | Weekly | Visual merchandising |
| Queue wait | Time from joining the queue to reaching the counter | Counter camera, vision model | Daily | Store manager |
| Time to approach | Time from entry to the first advisor contact | Floor cameras, vision model | Weekly | Store manager |
| Unattended visit rate | Visits with no advisor approach before exit | Floor cameras, vision model | Weekly | Sales head, store manager |
| Shelf interaction rate | Visitors in a zone who handled a product | Shelf cameras, vision model | Weekly | Merchandising |
What the camera cannot tell you
Video sees the approach and not the conversation. It can report that an advisor reached the customer in ninety seconds and that the customer left after six minutes without a bill; it cannot say whether the customer asked for a model the store did not carry, heard the wrong scheme, or was never asked what they came for. The event is recorded and its reason is not.
Joining camera events to consented conversations is the natural next step, so that an unattended visit, a long wait or a walk-away has the words that explain it. Borentis captures the conversation side today on the advisor's phone; ShopperDNA, which joins camera events to those conversations, is on the roadmap.
Go deeper
This page is an orientation. Two longer guides on this site cover the subject properly: what retail video analytics is, with the use cases and the limits, at /learn/what-is-retail-video-analytics/, and how to use existing CCTV for retail analytics, including camera placement, edge versus cloud processing and consent, at /learn/how-to-use-cctv-for-retail-analytics/.
Frequently asked questions
What is retail video analytics?
Retail video analytics applies computer vision to CCTV footage to produce counts and events, entries, paths, dwell, queues, advisor approaches and shelf interactions, instead of hours of recording that nobody watches.
Can existing CCTV cameras be used for retail analytics?
Usually yes, if the camera angle covers the entrance or zone of interest and the recorder or an edge device can run the vision model. Older analogue cameras and poor angles are the common blockers.
Is video analytics in stores legal in India?
CCTV for security is common practice with signage. Analytics that counts and tracks movement without identifying individuals sits within that practice; anything that identifies a person, such as face recognition, is personal data under the DPDP Act and needs a lawful basis. Most retail use cases do not need identification.
Related reading
- What is retail video analytics
- How to use CCTV for retail analytics
- Retail analytics: the full guide
- Footfall analytics
Where Borentis applies this
- Walk-in Recovery: The customer who left is still yours.
- Execution Scorecards: See the floor before the P&L does.
Borentis is the Agentic Operating System for Customer Interactions, built for Indian retail floors: consented one-tap capture on the advisor's phone, every conversation scored against your playbook with the evidence behind every number, leads created when a number is heard, and coaching from your own best conversations.