Small Retailer
Understand the Moment
Move beyond movement analytics. VLM interprets behaviour and context, transforming observed activity into defined, verifiable metrics that reflect real customer engagement and service interactions.
Why This Matters
Traditional analytics measure presence and duration. VLM introduces interpretation.
Distinguishes between types of behaviour within the same space.
Allowing identical movements to be understood differently based on context, posture, and interaction patterns.
Contextual Metrics
VLM converts observed activity into structured classifications. A person in front of a shelf is no longer a dwell event, but a defined behaviour-ranging from brief attention to active product evaluation. The same applies to staff interaction, service flow, and completion stages, forming measurable operational indicators.
Contextual Categorisation
VLM refines what a person is doing within a detected zone or interaction:
- Shelf zone: glance, browse, active product interest
- Staff interaction: asking, being served, upsell interaction
- Checkout: payment, completion, handover
This turns a simple “20 seconds in front of shelf” into a defined behaviour category.
Built-in Validations
The same mechanism verifies whether an observed action is genuinely meaningful:
- Distinguishes shopper engagement vs staff restocking
- Filters out idle or irrelevant presence
- Confirms whether interaction actually occurred
This removes ambiguity from traditional metrics and increases trust.
Same Moment, Different Meaning
A single location can produce multiple outcomes. A stationary person may be distracted, browsing, or engaged in product evaluation. VLM resolves this ambiguity by interpreting behaviour rather than relying on duration thresholds, enabling consistent categorisation across varying real-world conditions.
Phone Distraction
VLM Output: Person stationary, attention directed to mobile device. No product interaction.
Not Engaged
Casual Glance
VLM Output: Brief visual attention towards products, no hand interaction
Lvl1:Light engagement
Active Browsing
VLM Output: Sustained visual focus on products with exploratory behaviour.
Lvl2: Browsing
Deep Product Interaction
VLM Output: Direct product handling and evaluation behaviour.
Lvl3: High Engagement
Talking to Friend
VLM Output: Conversation detected, not related to product interaction
Not Engaged
Talking About Product
VLM Output: Shared attention towards product with discussion behaviour.
Lvl4: Staff Engagement
AI-generated illustrations only; no real customer data, surveillance footage, or personal information is used or captured.
Retail Applications
Across store environments, VLM enables consistent interpretation of key activities. Product engagement, staff interaction, and checkout processes can be measured as structured behaviours rather than inferred events, providing a clearer view of customer journey and operational execution.
Shelf Engagement
Staff Interaction
Checkout Completion
Verifiable Outcomes
Each classification is derived from observable behaviour and can be reviewed as a discrete event. This enables validation workflows where outcomes are inspected and refined, ensuring that metrics remain aligned with actual store activity rather than assumed patterns.
Operational Definition
Behaviour categories are not fixed. Retailers define what constitutes meaningful engagement or service quality within their own context. This allows consistent measurement across different formats, from high-touch retail environments to transaction-focused stores.
Built for Practice
The system is designed to operate within defined conditions, focusing on meaningful interactions rather than continuous interpretation. This ensures that analysis remains relevant, controlled, and aligned with operational use rather than theoretical modelling.