StationPro playbook

Gas station camera integration: tying CCTV events to cash and lottery.
Cameras show what happened; back-office data shows the dollar exposure. The integration pattern that links a flagged variance event to the matching CCTV clip, Verkada, Solink, March Networks, Hanwha API options.
What does camera integration actually do?
Without integration: the camera records continuously, the back office tracks variance, and the operator does the cross-reference manually. A $-130 cash variance on Tuesday's overnight shift means the operator scrubs through the camera footage looking for the moment something happened. That's a 30–60 minute process per event, so most variance events never get reviewed.
With integration: the variance event automatically links to the camera clip at that timestamp. Operator clicks the flag, sees the moment in 30 seconds, makes a decision (training, process error, investigate further, dismiss). The investigation cost drops from 30+ minutes to ~30 seconds, and the rate of investigated events goes up accordingly.
The major camera systems for gas stations
Verkada
Cloud-native security camera platform with a documented REST API. Most modern integration path. Camera footage accessible via API by camera ID + timestamp. Used widely at independent and chain c-stores; relatively expensive per-camera but no NVR/DVR maintenance.
Solink
Specifically designed for retail/c-store loss prevention. Integrates with major POS systems out of the box; the API is mature. The integration with back-office software is more turnkey than most other camera vendors.
March Networks
Enterprise-grade NVR system used in larger c-store chains and quick-serve restaurants. API exists but integration requires custom work. Common at branded gas stations (Shell, BP) where the brand specifies the camera vendor.
Hanwha (Samsung Wisenet)
Mid-market NVR + camera system common at independent stations. SDK available; integration is custom work, typically 1–2 weeks of engineering.
Legacy DVR systems
Older HiKvision, Dahua, Lorex, Swann DVR systems generally don't expose APIs that back-office software can consume cleanly. Path forward is usually a cloud-camera upgrade if loss attribution via camera is high priority.
How camera + back office integration works
1. Event detection in the back office
Cash variance over warning band, lottery serial gap, void burst, after-hours register open. The back-office system flags the event with a timestamp, store, and (where applicable) the responsible clerk.
2. Camera API call by timestamp
For each flagged event, the back-office system calls the camera API: "Give me the camera clip for store #3, register camera, 11:42:00 PM – 11:43:30 PM Tuesday." The API returns a streaming URL or downloadable clip.
3. Auto-link in the alert UI
The flagged event in the alert queue shows a thumbnail or play button next to the variance amount. Click to watch the clip inline. No tab-switching, no separate login.
4. Owner review and decision
Watch the 30–90 second clip. Make the decision: training opportunity, process error, investigation needed, or dismiss. The decision writes to the audit row alongside the original event.
5. Retention and privacy
Camera clips stay in the camera vendor's retention (typically 14–30 days at standard tier, longer at higher tiers). Back-office system stores the link + metadata but not the video itself, keeps the storage manageable and respects camera-vendor data ownership.
What camera-event types matter
Modern cameras detect multiple event types. The operator-relevant ones at a gas station:
- Motion in restricted areas. Office, safe room, after-hours store. Pairs with off-hours cash variance investigation.
- Register open / cash drawer open.Especially important during overnight shifts. Drawer opens that don't correspond to a transaction are a high-attribution signal.
- After-hours activity. Anyone in the store outside business hours. Pairs with safe-drop discrepancies.
- Suspicious behavior detection.AI-driven behavior flags (loitering, items concealed, repeated entry). Newer cameras only.
- Specific incident detection.Falls, crowd formation, theft suspected. Less common; depends on camera vendor's AI layer.
Example: a flagged variance with clip review
MORNING BRIEF. Store #3, Tuesday 7 AM
ATTENTION (1)
EOD variance overnight shift: $-87
Clerk: M.J.
Reason code: "no explanation logged"
Camera link: [▶ play] (Verkada cam-register-01, 23:42:15)
Loss Radar correlation: void burst at same window
+ lottery serial gap pack #B-024
Click play → 38-second clip:
11:42:15, clerk rings $13 Marlboro Box at $1.30 (decimal slip)
11:42:38, customer pays cash, takes pack
11:42:50, clerk voids the $1.30 transaction
11:42:54, clerk pockets $11.70
Decision: investigation needed; pull HR file, schedule
conversation before next overnight shift.
Audit row updated: "camera-confirmed cash-pocketing pattern,
HR escalation in progress"Privacy and labor-law considerations
Camera installation and monitoring at a workplace triggers state-specific labor-law requirements:
- Display signage. Most states require visible signs notifying employees and customers that the premises are monitored.
- Employee notification. Some states (Illinois, Connecticut, New York) require employee notification or consent for workplace recording. Document in the employee handbook.
- Audio recording is separate. Most retail security cameras record video only. Audio recording has stricter wiretapping/eavesdropping rules, two-party-consent states (California, Illinois, etc.) make audio recording materially more complex.
- Restroom and break-room exclusions.Federal law and most state laws prohibit cameras in areas with reasonable expectation of privacy.
For loss attribution, video-only cameras at the register, forecourt, lottery counter, and back-office cover all useful angles without triggering audio-recording complications.
Integration timeline and effort
| Camera vendor | API maturity | Typical integration time |
|---|---|---|
| Verkada | Documented REST API | Days; pre-built integrations |
| Solink | Mature integration platform | Days; turnkey for retail use case |
| March Networks | SDK available | 1–2 weeks custom work |
| Hanwha (Wisenet) | SDK available | 1–2 weeks custom work |
| Legacy DVR (HiKvision, Dahua, etc.) | Generally no API | Not feasible without hardware upgrade |
Frequently asked questions
What does camera integration do for a gas station back office?
Which camera systems integrate with back-office software?
Do I need cameras to use loss-attribution software?
How much storage does camera integration use?
What's required to legally record at a gas station?
Can I add camera integration to an existing camera system?
What camera-event types are most useful for loss attribution?
Sources & methodology
This playbook draws on operator workflows observed in StationPro pilot stations and on anonymized product data from live pilot tenants. Figures are illustrative examples, not promises about your stores. Procedures were reviewed against the workflows of the StationPro operator team before publication. Questions or corrections: talk to the team.
StationPro Editorial
The operator team behind StationPro. We write the procedures we ship: every playbook comes from real close, reconciliation, and loss-attribution workflows in pilot stations.
Keep reading
EOD in 5 minutes.
A teardown of the daily-close process at a typical independent gas station, what takes 45 minutes today, why it takes that long, and how the same close can be done in five minutes without cutting corners.
Invoice OCR vs. manual entry, in dollars.
A working calculation for the labor savings, error reduction, and price-leak capture when you stop typing vendor invoices and start scanning them.
Gas station exception reporting: how owners find the stores that need attention.
Exception reporting tags every event against the store's own baseline and surfaces only what's anomalous. Cash, lottery, fuel margin, inventory, expenses, deposits, explained with the daily review workflow that compresses 400 data points into 5–15 actionable flags.
