Employee Theft in Indian Factories: What Cameras Catch (and What They Don't)
Pilferage is money walking out of your gate — a few kilos of brass here, a "short" dispatch there, scrap that never gets weighed. Cameras won't grab the thief's hand, but pointed at the right five spots they deter casual theft, create a timestamped evidence trail, and flag the process breaches — unlogged movement, after-hours access, tailgating — that pilferage always leaves behind. Here's where to aim them and what actually holds up.
Most owners find out about internal theft the way you'd least want to: a stock-take that won't reconcile, a customer complaint about a short load, or a whisper from a loyal supervisor. By then the money is gone and the trail is cold. The question isn't "are my people honest" — it's "where is my floor blind, and what would a camera there actually let me prove."
Key points
- Where, not whether. Pilferage clusters at a handful of choke points: the security gate, the raw-material store, the dispatch/loading bay, the tool crib, and the scrap yard. Cover those first.
- Cameras deter and document — they don't physically stop theft. A camera is a witness, not a guard. It changes behaviour and builds a case; it doesn't grab anyone's hand.
- The real value is the evidence trail and the process breach. After-hours access, a truck that leaves before it's logged, a person tailgating through the gate, a store opened at 2 a.m. — these are visible even when the theft itself isn't.
- In India, footage is only useful if it's admissible. Since 1 July 2024 the Bharatiya Sakshya Adhiniyam governs electronic evidence; a certificate (the old Section 65B, now Section 63) from a responsible person is a condition precedent for CCTV to hold up in court.
- Be honest about limits. A camera can't read every pocket, can't weigh a truck, and can't prove intent on its own. Pair it with logs, weighbridge data, and gate registers.
The five hotspots — and what a camera can honestly do at each
Below is the map we walk owners through. Read the caveat column as carefully as the "catch" column — over-promising here is how CCTV projects lose trust.
| Theft type | Where it happens | Can a camera catch it? | Honest caveat |
|---|---|---|---|
| Goods out the gate (finished product, spares in a bag/tiffin) | Security post, main and side gates, wicket doors | Deters + documents. Records who left, when, carrying what; flags tailgating and use of unmonitored side exits | Won't see inside a closed bag or pocket; needs good lighting and a clear "search line." Evidence is behaviour + timing, not X-ray |
| Raw-material pilferage (brass, copper, resin, yarn, packaging) | Raw-material store, issue counter, WIP staging | Catches process breach. After-hours store entry, issue without a slip, quantities that don't match the log | Can't count kilos by eye. Value comes from pairing footage with the issue register and stock system |
| Short dispatch / loading fraud | Loading bay, weighbridge, dock doors | Strong. Timestamps every truck in/out, links a load to a gate pass, catches loading before paperwork or after hours | A camera doesn't weigh — weighbridge tampering is a separate control. Use footage to corroborate weighbridge and challan data |
| Tool & consumable walk-off (bits, gauges, PPE, spares) | Tool crib, maintenance store | Deters + logs issue/return. Flags who accessed the crib and whether items came back | High-churn small items are hard to itemise on video; treat as deterrent + spot-check aid, not an inventory system |
| Scrap diversion (off-cuts, rejects, metal scrap sold on the side) | Scrap yard, scrap-out gate, back boundary | Catches unlogged movement. Scrap leaving outside scheduled lifts, unrecorded vehicles at the boundary | Scrap value is easy to under-declare; camera proves movement, weigh-and-log proves quantity. You need both |
Notice the pattern: a camera is strongest where theft requires movement through a fixed point and weakest where it hides inside a closed container or in a number on a slip. That's why dispatch and gate are the highest-ROI installs, and why the scrap yard and store need a camera plus a logging discipline to pay off.
What the numbers say (and what we won't overstate)
We're deliberately careful here, because internal-theft statistics get repeated loosely. What's reasonably grounded:
- In the retail sector, India has repeatedly ranked among the highest in the world for shrinkage, and Indian retail chains report shrinkage rising as sales volumes grow — one listed retailer disclosed shrinkage climbing to 0.41% of sales in FY24 from 0.22% the year before. Manufacturing isn't retail, but the direction of travel — more throughput, more leakage — rings true on a busy floor.
- In retail-barometer data, employees are consistently the second-largest source of shrinkage after shoplifters — one India read put employee theft at roughly a quarter of losses. We flag this as a retail figure; we have not found a clean, comparable factory-floor number, so treat any "X% of factory losses are internal" claim with suspicion.
- Where hard cases exist, they're eye-opening: Indian police have busted weighbridge-tampering rings — including a 2026 Ahmedabad case pegged at up to ₹100 crore — where chips and remote controls shaved 5–10 tonnes off a truck's recorded weight at the press of a button. That's dispatch fraud at industrial scale, and it's exactly the loading-bay blind spot a timestamped camera + weighbridge log is meant to close.
The honest takeaway: you don't need a scary national percentage to justify action. Take your own annual raw-material and finished-goods throughput, assume even a fraction of a percent leaks at your weakest choke point, and the rupee figure usually dwarfs the cost of covering it. Our theft and shrinkage monitoring overview walks through sizing that loss for your own floor.
Making the footage actually count in India
Catching it on camera is step one. Making it stick — in an internal enquiry, a full-and-final settlement, a police complaint, or a labour proceeding — is step two, and it's where most factories fall down.
- Admissibility. Under the Bharatiya Sakshya Adhiniyam, 2023 (in force since 1 July 2024, replacing the Evidence Act's Section 65B), CCTV is secondary electronic evidence. A certificate from a responsible person, confirming the system was working and the recording is genuine, is a condition precedent — courts have rejected footage produced without it, and rejected certificates cooked up years later as an "afterthought."
- Chain of custody. Recording must be continuous and in the ordinary course of business, the device shown to be functioning, and the export handled cleanly. Retention matters: if your DVR loops over 7 days, a theft found at month-end is already gone.
- Use it as corroboration. Indian courts generally treat CCTV as corroborative — decisive alongside a gate register, an issue slip, a weighbridge printout, or a statement, rarely on its own. Design your controls so the camera confirms the paper trail.
This is where "a second brain for the owner" earns its keep: the value isn't watching screens, it's the system quietly flagging the process breach — store opened after hours, truck out before challan, scrap gate used off-schedule — and holding the clip with its timestamp so you have something real to act on the same day, not a cold trail next quarter.
Camera plan vs. more guards
Owners often ask whether the answer is simply more men at the gate. Guards deter, but they're a recurring cost, they can be complicit, and they don't leave a searchable record. A camera doesn't get tired, doesn't take a cut, and its footage can be reviewed after the fact. In practice the strong setup is fewer, better-placed guards backed by cameras at the five hotspots — we break down the economics in our security guard vs. AI camera cost comparison, and you can put your own numbers in the guard ROI calculator.
FAQ
Will a camera stop my workers from stealing? No — and any vendor who promises that is overselling. A camera deters casual theft and gives you a timestamped record to act on. Physically stopping theft still needs gate discipline, logs, and controls. The camera is your witness, not your bouncer.
Can I use the footage to take action against an employee? Yes, if it's handled right. For an internal enquiry or a police complaint in India, keep continuous recordings, preserve the clip before it loops, and get a certificate from a responsible person per the Bharatiya Sakshya Adhiniyam. Treat CCTV as corroboration alongside your registers and slips.
Where should I put my first cameras if budget is tight? Dispatch/loading bay and the main gate first — that's where the largest single losses move through a fixed point. Then raw-material store, scrap-out, and tool crib.
What about weighbridge and scrap fraud — can cameras solve that? They cover half of it. A camera proves movement and timing (which truck, when, was it logged); it can't weigh. Pair it with weighbridge data and a scrap-lift log to prove quantity.
How long should I keep recordings? Long enough to survive a month-end stock-take — 30 days is a sensible floor. Many thefts surface only when the numbers don't reconcile, well after a 7-day loop has erased the evidence.
JSON-LD note for publishing: emit Article schema (headline, author, datePublished, publisher = Mama) and a nested FAQPage built from the five Q&As above, so the answers are eligible for rich results and AI-answer citation.
Want this mapped to your floor? Request a camera plan and we'll mark the five hotspots for your layout, with a PoE and mounting budget.