The shoplifter the camera saw and the staff never did
A man slides two bottles of spirits into a backpack in aisle nine. The ceiling camera records the whole thing in full colour. Nobody in the store sees it until three weeks later, when a stock count reveals the missing items and a loss prevention analyst scrubs through hours of footage to find the moment. By then the goods are gone, the man is gone, and the footage is nothing more than evidence for a police report that will almost certainly go nowhere. This is not unusual. This is how loss prevention works across most of the retail sector. It plays out on camera, in stores with full CCTV coverage, at a scale of roughly $112 billion per year.
Why loss prevention is always too late
The pattern is built into the process. A specialist team reviews footage after a problem has already been found. The review happens after the fact. The goods are already gone. The moment to step in has passed. The cameras a retailer has paid to install across its entire estate end up as a tool for explaining what happened on a given shift, not for changing what is happening on the shop floor right now.
The industry has tried to fix this before. Full vision systems built around proprietary cameras, closed software platforms, dedicated hardware. Those programmes typically require six-figure budgets per chain, take months to deploy across even a single region, and stall before reaching the smaller stores where theft is most concentrated. A handful of flagship sites work as intended. The rest of the camera estate keeps running the same way it has for twenty years. That combination, a process that only looks backward, and the repeated failure of expensive replacements, is why retail shrinkage has stayed unsolved at scale, despite cameras being everywhere.
What changes when cameras act in real time
Manako connects directly to the cameras a retailer already owns. No new hardware. No engineers. No code. A Vision Agent runs locally at each store, acting on the specific situations a loss prevention team has defined: concealment in high-shrinkage aisles, exits without a checkout event, return-fraud patterns at the service desk, repeat visits by previously flagged individuals across the chain.
The difference is in what happens next.
When a Vision Agent identifies a concealment event, it does not add another alert to the queue of notifications staff learned to ignore years ago. It writes a structured event directly into the store's existing systems, showing up as a prompt on the handheld device the floor associate is already carrying. Staff act on it immediately, before the loss has actually happened.
When the loss does happen despite the intervention, the event already contains everything a loss prevention team would previously have spent hours piecing together: timestamp, camera, aisle, visual signature, clean clip ready for a case file. The full audit trail builds itself from the moment Manako goes live. Acting, not reviewing
The deployment runs on cameras and computing hardware the retailer already owns. No new capital spend for procurement to justify. Footage never leaves the premises, data privacy is handled by design, not by contract.
Manako Vision Agents are sub-50MB, running on standard hardware at a fraction of the cost of dedicated loss prevention systems. Take a chain operating five hundred stores, losing an average of forty thousand dollars per store per year to preventable theft. A twenty percent reduction across the estate covers the cost of the rollout within the first quarter. The savings grow from there as the loss prevention team builds out additional Vision Agents for additional types of incident. Compliance and audit readiness are built in from day one.
The $112 billion question
Loss prevention has worked the same way since the first CCTV camera was installed in a store. The cameras have always seen the shoplifter. The question was never whether the footage existed. The question is whether anything is going to act on what the cameras see, in the brief window where acting would still change the outcome.
That window has been opening and closing, on camera, $112 billion a year. It does not need to stay that way.
Tell Manako what you need. It does the rest.
Join the waitlist to be the first to build with Manako, or get in touch to discuss an enterprise deployment across your operation.