Question 1

What is a People Counting system and how does it work?

Accepted Answer

A People Counting system, or software, uses physical sensors (typically video cameras or infrared beams) to record how many people enter and exit a physical space. Modern software like Flame Analytics applies AI video analytics to existing cameras, eliminating the need for dedicated hardware. Flame's Hypersensor technology processes video streams in real time, detecting and counting people through silhouette recognition without identifying faces or using biometric data. Flame reaches over 99% accuracy in physical spaces, stores, and shopping malls across 7 countries. Unlike traditional infrared counters that only measure door crossings, AI video counting works across complete zones, distinguishing employees from visitors.

Question 2

How accurate is AI People Counting compared to manual counting?

Accepted Answer

AI-based People Counting like Flame's reaches over 99% accuracy in controlled environments, far surpassing manual counting, which typically delivers 75-85% accuracy due to human fatigue and distraction. Flame's Hypersensor uses deep learning models trained on millions of real-world scenarios to maintain accuracy even in high-density environments such as shopping mall corridors or store entrances during peak hours. The system self-calibrates using historical patterns and handles complex scenarios (groups walking together, people with large bags, wheelchair users) that cause traditional infrared or thermal sensors to undercount by 15-30%. Accuracy is validated through field audits for enterprise clients, including leading retailers, shopping mall operators, and organizations managing physical spaces.

Question 3

How much does a People Counting system for a store cost?

Accepted Answer

Total cost depends on three factors: number of entrances to monitor, existing camera infrastructure, and analytics modules selected. Flame can work with the client's existing cameras, though it will depend on camera quality and proper placement and calibration, eliminating the per-entrance cost of dedicated sensors. Software licenses follow a SaaS model with monthly or annual subscriptions. For a single location with 2-4 cameras, expect EUR 150-400/month, including People Counting, basic traffic analytics, and dashboard access. Organizations with multiple locations—from a large fuel retailer (over 1,000 stations) to a network of physical spaces—benefit from volume pricing. ROI typically materializes in 3-6 months through staffing optimization alone, reducing labor costs by 8-15%.

Question 4

What is the ROI of implementing People Counting?

Accepted Answer

Organizations using Flame's People Counting—whether they manage physical spaces, stores, or shopping malls—achieve measurable ROI in 90 days across three areas. First, conversion rate visibility: knowing the real ratio of visitors who buy reveals whether low sales are due to lack of traffic or poor in-store experience. Second, staffing optimization: aligning employee shifts with real traffic patterns reduces labor costs by 8-15% and improves visitor service during peak hours. Third, marketing attribution: measuring footfall lift after campaigns demonstrates which promotions truly generate visits. Flame clients from the portfolio of leading sector operators with 10 shopping malls increased their tenants' conversion rates by an average of 12% in the first year.

Question 5

Is People Counting GDPR-compliant?

Accepted Answer

Yes, when implemented correctly. Flame's People Counting is privacy by design: the system counts anonymous silhouettes moving through the space without capturing, storing, or processing biometric data. No facial recognition, no individual tracking, no personal data collection. Data is aggregated into statistical counts—147 visitors between 10:00 and 11:00—making it impossible to identify any individual. Flame holds ISO 27001 certification and processes all data on EU servers. The system generates only aggregated metrics (hourly counts, occupancy levels by zone, average dwell times), never individual profiles. This approach exceeds GDPR requirements and applies equally to physical spaces, stores, and shopping malls.

Question 6

Does it work with my existing video surveillance cameras?

Accepted Answer

Yes, and this is Flame's key differentiator. Flame's Hypersensor technology connects to existing IP cameras via RTSP protocol, transforming video surveillance infrastructure into business intelligence with no new hardware. Compatible with major camera brands including Axis, Hikvision, Dahua, Bosch, and Hanwha. Requirements are minimal: cameras need at least 720p resolution and a top-down or angled-from-above mounting position. Flame's technical team performs a free camera audit to confirm compatibility before deployment. In a large fuel retailer's deployment across more than 1,000 service stations, Flame connected to the existing camera network without any hardware purchases, reducing deployment cost by 60%. The same approach applies to physical spaces, stores, and shopping malls.

Question 7

What's the difference between real-time and historical data?

Accepted Answer

Real-time data shows current occupancy and traffic flow as it happens, essential for immediate decisions like opening additional service points or activating capacity management protocols in physical spaces. Historical data reveals patterns across weeks, months, and years, enabling strategic decisions about opening hours, seasonal staffing, and layout changes. Flame provides both through a unified dashboard. Real-time data refreshes every 15 seconds with configurable threshold alerts. Historical analytics includes year-over-year comparisons, day-of-week patterns, and weather correlation analysis. Physical spaces use real-time data for operations, while shopping malls and stores rely on historical data for quarterly business reviews and lease negotiations.

Question 8

How long does installation take?

Accepted Answer

For a single location with existing cameras, Flame typically delivers live counting data within 5-7 business days from contract signing. The process: Days 1-2, camera audit and RTSP connection setup (remote); Days 3-4, AI model calibration for your specific environment; Days 5-7, dashboard configuration and team training. No on-site hardware installation is required when using existing cameras. For multi-site deployments, Flame implements in phases: a large retailer's multi-store rollout followed a 4-week phased approach, and the same methodology applies to any network of physical spaces. Enterprise clients receive a dedicated project manager and custom API integration support.

People Counting

Benefits of People Counting

Understand your real footfall

Efficient staffing and operations

Compare locations easily

Boost conversion and profitability

Beyond counting: intelligence on your footfall

The metrics that change decisions

Total and hourly footfall

Visit-to-sale conversion rate

Weekly patterns and peak hours

Store-to-store benchmark

Average time in store

Staff-to-traffic ratio

Solutions for any Industry

Shopping malls

Retail

Public venues

Hospitality

The best brands talk about us

Frequently asked questions

Request a demo