Concessions and Retail Loss Prevention at the Point of Sale
The Shrink Isn’t Walking Out the Door. It’s Ringing Through the Register.
Exception reports have been telling you which transactions look wrong for years. Nobody has the hours to watch the video. POS Track closes that loop.
Who This Is For
This is for the VP of Concessions or Retail Operations running many units with high seasonal turnover, and for the Loss Prevention lead who has the exception reports but not the headcount to work them.
You already know your shrink number. What you can’t do is attribute it, because that means pairing a flagged transaction with what physically happened at the counter. That pairing is manual and slow, so it rarely happens.
Where the Model Breaks
Register-side loss is the loss you can’t see. Sweethearting, under-ringing, void and refund manipulation, no-sale drawer opens and cash skimming happen in plain view of a camera nobody is watching, by people you trained.
Exception reporting produces leads, not cases. Your POS system flags anomalies faithfully. Turning a flag into a case means pulling video, finding the transaction and reviewing it, perhaps twenty minutes per lead against hundreds of leads a week.
Seasonal and event staffing defeats your controls. Concessions and event retail run on high-turnover, quickly trained crews. Your controls were built for tenured staff and are applied to people on their third shift.
Loss prevention can’t cover the unit count. One LP professional against dozens of points of sale across a venue or region means coverage is reactive, driven by whichever unit’s variance got bad enough to notice.
Queue abandonment is a revenue leak nobody owns. At a venue, a line longer than the crowd will tolerate is lost revenue in a fixed window. It sits between operations and security, so neither measures it.
An Operations Decision, Not a Security One
Register-side loss is recurring, budgeted as a cost of doing business, and can be larger than the security spend on the whole footprint. Size it from your own P&L and exception data, not category averages. That’s exactly what Phase 1 measures.
What NXTFusion OS™ Brings
NXTSensorHub: see everything. It connects your existing overhead and counter cameras with the POS transaction stream, so detection runs on infrastructure already installed at every unit.
POS Track: miss nothing. It pairs flagged transactions with what physically happened at the counter, turning exception-report leads into reviewable cases automatically.
Nexie: respond. The AI Operator prioritizes by exposure rather than volume, surfaces patterns across units and shifts, and produces the case record LP and HR work from.
Governance, Stated Up Front
AI Assist: Nexie detects, classifies and prioritizes, then drafts the incident record with the evidence attached. Your operator decides.
AI Orchestration: Nexie sequences the response. It repositions the asset, cues the nearest camera, dispatches the responder, and opens and time-stamps the log.
Red lines: No autonomous use of force. No autonomous contact with a person. No identity determination without written policy and counsel review.
One Shift at One Register
Transaction: A void, refund, no-sale or discount fires an exception. Today this is where the trail stops, for lack of hours to follow it.
Pairing: The transaction is matched to what physically happened at the counter. The twenty-minute manual step that made exception reports unworkable disappears.
Pattern: Repetition across shifts, units and operators is surfaced as a pattern. Single incidents are noise; the pattern is the case.
Case: A reviewable package reaches LP with evidence attached, ready to act on rather than another line on a report nobody finishes.
Deterrence: Once staff understand exceptions are actually reviewed, behavior changes before any case is made.
Built for Venues
At a stadium or arena, concessions, parking and security often report to the same operations executive. Where NXTFusion OS™ already secures the exterior estate, POS Track adds a second line of value on the same relationship, and the concessions P&L often approves faster than the security budget. See how the platform covers the lots in the stadium and arena use case, and how RevTrack protects parking revenue.
Where It Already Runs
Snapdragon Stadium and Petco Park: High-volume, high-turnover concessions under compressed event windows, the hardest control environment in food and beverage retail.
Pechanga Arena: Live concert and event operations with surge staffing across many points of sale in a single evening.
Sutter Health Park: Multi-zone venue operations with distinct retail and concessions setups by event type.
San Diego International Airport: A high-throughput, credentialed-staff retail environment with continuous operation and formal audit expectations.
What a Phased Deployment Looks Like
Phase 1: One venue or region, 60 days. Baseline attributable register-side loss against your existing exception data. Measure, don’t project.
Phase 2: Full unit coverage. Case generation is integrated into the LP workflow, with patterns surfaced across units, shifts and operators.
Phase 3: Control redesign. Rebuild register policy and training around what the data actually shows.
Frequently Asked Questions
Do we have to replace our POS exception reporting?
No. POS Track works with the exceptions your POS system already flags and pairs each one with what physically happened at the counter, using your existing cameras.
What kinds of register loss does it address?
The source use case covers sweethearting, under-ringing, void and refund manipulation, no-sale drawer opens and cash skimming.
Who reviews the cases?
Nexie prioritizes by exposure and drafts the case record with evidence attached. Your LP and HR teams review it and decide.
What does a first deployment look like?
Phase 1 is one venue or region for 60 days, baselining attributable register-side loss against your existing exception data.
Next Step
Run sixty days at one venue against your existing exception data. Either way, you’ll know what your attributable register-side loss really is.