Straight answer: a jewellery shop needs four specific camera positions covering the counter, the strongroom entry, the shop entrance and the street outside — not simply more cameras. The counter camera, angled to see hands and the display tray, is what catches distraction theft. Cameras document a robbery in progress; they don't stop one. And staff-side risk is best handled through routine reconciliation applied to everyone, not through watching individuals.
After a scare — a snatch attempt, a near-miss, a robbery down the road — the usual reaction is to call the CCTV vendor and ask for more cameras at the door. That's understandable, and it's also the wrong first question. The door is rarely where jewellery retail loses money or gets hurt.
Where does jewellery theft actually happen inside the shop?
Mostly at the counter, through distraction rather than force. A customer asks to see several pieces at once — a common, genuinely innocent request most of the time — and while the salesperson's attention is split between trays, an accomplice swaps or pockets a piece. This works because the salesperson's eyes are on the customer's face, not on their hands or the tray. A wide camera at the entrance, thirty feet away, never sees this clearly enough to matter.
What should the counter camera actually show?
Hands and tray, not faces. A camera mounted to look down and across the counter — close enough to resolve individual pieces moving — is what actually settles a disputed count afterwards. If your current camera gives you a good headshot of every customer but a blurry wide shot of the display case, it's set up for the wrong problem.
What about the strongroom or vault?
This is where access control matters more than the camera itself. A camera at the strongroom door records who went in and when; it doesn't decide who's allowed to. Pair it with a log — even a simple register cross-checked against the footage — so that "who opened the safe on Tuesday at 2pm" has one answer, not three conflicting ones from memory.
Can cameras stop an armed robbery?
No, and any pitch that implies otherwise is overselling. Once someone is inside, armed, and past the door, a camera's job is to record for identification and insurance — it has no physical way to intervene. What actually slows a robbery down is a time-delay lock on the strongroom, a display case that can't be opened by one person alone under pressure, and a panic alarm that reaches the police or a response team quickly. Cameras belong alongside these controls, not instead of them.
Does the street outside matter?
Often more than people expect. Robberies are rarely spontaneous — someone frequently loiters near the shop or walks past repeatedly over several days, checking staff timing, the lock-up routine, and which side the alarm panel faces. A camera covering the street and the approach to the entrance, reviewed for that kind of repeat pattern, can flag the casing stage while there's still time to act — change a routine, alert the local police beat, or simply be more alert that week. Once the robbery itself starts, the same camera's role shifts from prevention to record-keeping.
How do you handle staff-side risk without accusing anyone?
Carefully, and as a routine applied to everyone rather than a hunt aimed at one person. Reconcile stock counts against transaction records on a fixed schedule. Keep the strongroom access log genuinely reviewed, not just recorded. Frame it, and mean it, as a control that clears honest staff fast when something doesn't add up — because in a small team, an unresolved discrepancy is corrosive to trust in both directions. Singling someone out based on a hunch does more damage to morale than the loss you're trying to prevent.
What's the honest limitation here?
Camera footage and access logs settle disputes about what happened. They don't replace the physical delays — locks, time-release safes, alarm response — that actually slow an armed intruder down, and they can't fully substitute for a staff member's judgement about who they're serving. A shop that adds analytics on top of good physical security gets real value; a shop that adds analytics instead of physical security is just recording its own loss in better resolution.
We build camera-based analytics for retail security including counter and entrance coverage, so factor that interest in when weighing this against other advice. It's not the right fit for a shop that hasn't yet fixed the basics — lighting, lock quality, a working panic button — since no amount of footage review compensates for those being absent.