What can cameras actually measure with business-decision confidence? Retailers deploying computer vision for store analytics hear promises about measuring everything: customer traffic flow, dwell time, demographic breakdown, shelf compliance, queue length, associate-customer interaction frequency, and heatmaps of every aisle. The technology can detect most of these signals. The question that matters — and that vendor demonstrations rarely address — is at what accuracy level and under what conditions. Store analytics CV must distinguish “detected” from “measured with business-decision confidence.” Most deployments conflate the two. Detecting that a person exists in frame 87% of the time is technically impressive. Making a planogram compliance decision based on 87% detection accuracy means 13% of shelves are misreported — a rate that renders the measurement operationally unreliable for automated restocking triggers. Analytics by confidence tier Measurement Typical accuracy Confidence tier Suitable for automated decisions? Entrance/exit counting 95–98% High Yes — traffic trend analysis, staffing models Queue length estimation 90–95% High Yes — dynamic checkout opening triggers Aisle-level traffic flow 85–92% Medium Directional trends only — not individual path tracking Dwell time per zone 80–90% Medium Relative comparison between zones — not absolute measurement Shelf out-of-stock detection 75–90% (varies with shelf density) Variable Only with domain-trained models on specific planograms Planogram compliance 70–85% (per-SKU) Low for automation Requires human review loop — not reliable for automated replenishment Demographic estimation 65–80% (age/gender) Low Aggregate trends only — individual-level inference is unreliable and ethically problematic Customer-associate interaction 60–80% Low Counting presence proximity — not quality of interaction The shelf compliance problem specifically Shelf compliance monitoring requires sub-SKU-level accuracy that generic object detection models do not achieve without domain-specific training. Distinguishing between a 500ml and 750ml bottle of the same brand, detecting a product faced backwards, or identifying a product placed in the wrong planogram position are tasks that require: Training data specific to the store’s actual product catalog Camera positions that provide sufficient angle and resolution per shelf Models retrained when seasonal product rotations change the planogram Lighting consistency or lighting-invariant model architectures In our retail CV deployments, shelf compliance detection achieves commercially useful accuracy (>90% per-facing) only when the model is trained on the specific store’s product range and camera geometry. Generic models from transfer learning achieve 70–80% — useful for research demonstrations, insufficient for automated restocking decisions. What actually drives retail CV ROI The measurements with highest confidence — entrance counting, queue length, zone traffic — are also the measurements with most direct operational impact. Staffing decisions based on traffic prediction, dynamic checkout lane opening based on queue detection, and A/B testing of store layout changes based on flow measurement all operate in the high-confidence tier. The full ROI analysis of computer vision in retail shows that returns concentrate in these high-confidence applications. Attempting to extract ROI from low-confidence measurements (demographic profiling, interaction quality scoring) typically produces negative returns because the measurement uncertainty exceeds the decision threshold. Deployment architecture for store analytics Production store analytics typically runs on edge devices (one per 4–8 cameras) with a central aggregation server per store and cloud-level fleet analytics. The architecture decision that matters most: where does the “confidence” assessment happen? Systems that report raw detections to a central dashboard (without per-store accuracy calibration) produce analytics that look precise but contain systematic biases that vary by store, camera position, and time of day. Per-store calibration — running a ground-truth comparison for each camera position during installation, then adjusting confidence thresholds per-zone — is what separates analytics that inform decisions from analytics that create false confidence.
