Beyond revenue — see what customers order together, how often they return, and how long they wait. The data to build loyalty, not just transactions.
Average basket value, top item combinations, and order patterns — per shift, per day, per week. Know your revenue per transaction, not just per cover.
A table of two ordering a starter each, a main each, and one dessert to share is a very different customer than a quick lunch solo. Basket composition makes those differences visible.
Top order combinations — this week
683 orders
Avg basket value this week: €37.20
Co-occurrence
68%
Lift
×3.2
Co-occurrence
61%
Lift
×2.8
Co-occurrence
54%
Lift
×2.4
Co-occurrence
47%
Lift
×2.1
Co-occurrence
38%
Lift
×1.9
Co-occurrence
31%
Lift
×1.6
When Espresso and Tiramisu appear together in 68% of orders, that's not coincidence — it's a menu placement and upsell opportunity your staff should know about.
Affinity pairs surface the strongest item combinations automatically. Use them to train your team, design combo offers, or restructure your menu sections.
What lift means
A lift of ×3.2 means customers who order Espresso are 3.2× more likely to also order Tiramisu than the average customer. The higher the lift, the stronger the signal.
Frequency data segments your customers by how often they return. Regulars spend more, order differently, and are your most cost-efficient customers to keep.
If 42% of your covers are first-time visitors and your returning rate is flat, your experience isn't converting. Frequency data makes that visible before it becomes a revenue problem.
Visit frequency — last 90 days
Service timing — avg this week
How long from seating to first order? From order to delivery? Ticket timing data reveals bottlenecks without cameras — surfaced automatically from your POS.
A slow Main order → Main served time flags a kitchen bottleneck. A long Seated → First order gap flags a floor staffing issue. The data tells you where the problem is before a customer complaint does.
Data source
Ticket timing is derived from POS order timestamps. No additional hardware required. Accuracy improves with full POS integration.
Dwelitics will use anonymised Vision Analytics data to surface patterns that POS data alone can't reveal: walkout detection, idle table detection, queue length estimation, and zone heatmaps.
Walkout detection
Identify customers who leave before ordering — and when it happens most.
Idle table detection
Flag tables that have been cleared but not reset, reducing turn time.
Zone heatmaps
Understand which areas of your venue drive the most activity.
Queue estimation
Detect queue build-up at the bar or entrance before it becomes a complaint.
No personal identification. No biometric data. No cloud video upload. Vision Analytics processes everything on a local edge device — privacy by design.
Customer Behavior Intelligence
Beyond revenue — the data to build loyalty,
not just transactions.
Basket, affinity, frequency, and timing — four lenses on the same customer. All derived from your existing POS data. No new hardware required to get started.
Book a personalised demo and we'll walk through your customer data setup.
Continue exploring