Quick Commerce Now Owns the Stock: Insuring Spoilage, Theft, Transit Loss and Licence Suspension After Hyderabad's Dark-Store Raids

Telangana Today reported on 30 September 2026 that Telangana's TG SAFE teams (ten teams of food safety officers working alongside police) had inspected 16 dark stores in Hyderabad run by Swiggy Instamart, Flipkart Instakart, Zepto and Dropx Logistics. Four stores had their FSSAI licences suspended with immediate effect: Instamart's Pet Basheerabad and Saidabad stores, Flipkart's Panjeshah Road store and Zepto's Ram Koti store. Officers drew 15 samples and issued 3 improvement notices and 5 show-cause notices.

One in four inspected stores lost its licence in a single drive. That ratio matters more than any single finding, because it suggests to operators and their insurers that a licence suspension is not a tail event for a dark-store estate. It is an operating risk, and it arrives at the same time as a structural change in who owns the stock sitting on those shelves.

Our June piece on dark-store food safety, recall and product liability set out why a suspended licence is a liability and recall problem, not a property one. This post picks up where that one stopped. It looks at the inventory itself: the expiry, damage, pilferage and transit losses that now sit on platform balance sheets, and which of them any policy will actually pay for.

Under a pure marketplace model, a seller owned the goods until a customer bought them. Expired yoghurt, a crushed carton or a missing phone charger was the seller's write-off, settled through commercial terms with the platform. Once the platform buys the stock and holds it in its own dark stores, those losses become the platform's own cost of goods.

Blinkit has already made that move, and its numbers show the scale. Its Q1 FY27 shareholder letter, as reported by Apparel Resources on 22 July 2026, put losses from expired inventory, damaged goods, items lost in movement and theft at about 1.8% of net order value. On NOV of Rs 17,132 crore, that is roughly Rs 308 crore in a single quarter, close to three times Blinkit's Rs 102 crore adjusted EBITDA for the same period.

Swiggy is heading the same way. MediaNama reported that Swiggy shareholders approved a 49.5% foreign-ownership cap at the 18 August 2026 AGM, which opens the path for Instamart to shift to an inventory model. Swiggy expects that shift within two to four quarters.

The useful exercise for a risk manager is to break the shrinkage line into causes and test each against an insurance trigger. Indian property and crime wordings pay for sudden, accidental, identifiable events. They do not pay for the predictable cost of running a perishable, high-velocity retail operation.

• Expired inventory: stock that passes its best-before date because demand forecasting missed. This is a trading loss. No standard wording covers goods that simply aged on the shelf, and deterioration-of-stock extensions require a defined trigger such as refrigeration breakdown or power failure.
• Damaged goods in store: crushed packs, leaking bottles, breakage during picking. Small, frequent handling damage sits below any sensible deductible. Larger events (a racking collapse, a fire, a flood, a cold-room failure) are insurable material damage.
• Items lost in movement: stock that leaves a mother warehouse and does not arrive at the dark store, or arrives short. Insurable in part under transit or stock-throughput cover, provided the loss can be tied to a specific consignment.
• Theft: split between forcible external theft (burglary), employee or rider dishonesty (fidelity) and unexplained shortage found at stock count. The last of these is excluded almost everywhere.

On most accounts, the insurable share of the 1.8% is the minority. The bulk is expiry and small-ticket handling damage, which belong in operating controls and pricing, not in a claims file.

Once the platform owns inventory from the brand's dispatch point to the customer's doorstep, the stock passes through several locations in a day: the brand's factory or distributor, a mother hub, inter-city trucks, dark stores and the last-mile rider. Insuring that chain with a separate fire policy per location plus a separate transit policy creates gaps at every handover and a sum insured problem at every site, because dark-store stock levels swing daily.

A stock-throughput policy combines transit and storage cover on one form, usually on a declaration or annual-turnover basis. For a quick-commerce operator it answers three practical problems:

1. Fluctuating values: cover follows the stock rather than a fixed per-location limit, which reduces the risk of the average clause cutting a claim at an under-declared node.
2. Handover gaps: the policy attaches when the platform takes title, not when goods reach a named address.
3. Short-delivery evidence: losses in movement are claimable when they can be tied to a consignment note and a receiving record at the dark store.

What it will not do is pay for shortages discovered at a periodic stock count with no identifiable transit event. Insurers will want scan-level receiving data at each dark store before they accept that a loss happened in transit rather than on the shelf. For a related structure used by importers, see our note on FTWZ stock-throughput cover.

Rider-carried goods are the hardest leg. Many marine and transit wordings exclude goods in the custody of two-wheeler riders or limit cover to declared vehicles. If last-mile loss is material, it needs to be placed on purpose, not assumed.

Theft in the shrinkage figure has two very different sources, and they go to two different policies.

External theft

Burglary insurance on Indian wordings responds to theft following forcible and violent entry into, or exit from, the premises. A dark store running round the clock with staff present is rarely broken into in the classic sense. Shoplifting by walk-ins is uncommon because dark stores are closed to the public, but theft by outsiders posing as riders or delivery staff can happen. Ask whether the wording has been extended to cover theft without forcible entry and hold-up, and at what sub-limit.

Employee and rider dishonesty

Pilferage by pickers, store staff and delivery partners is the larger theft exposure for most operators. That is a fidelity guarantee or commercial crime exposure, not a property one. The key wording points are:

• whether gig riders engaged through a third party count as "employees" under the policy definition;
• whether the insurer requires the dishonest individual to be identified, which is difficult across a high-churn workforce;
• the inventory-shortage exclusion, which bars claims proved only by comparing book stock with a physical count.

Our earlier guide to crime insurance for quick-commerce operators covers the wording options in more depth.

Hyderabad shows the other side of owning the stock. When an FSSAI licence is suspended, the platform loses trading at that node, may have to destroy or withdraw condemned goods, and faces a higher chance that a consumer claims for illness from a product it sold. The stock is usually physically intact, so the property programme does not engage.

Three covers respond to different parts of that event:

1. Product liability pays third-party bodily injury and property damage claims arising from products the platform sold. Under an inventory model the platform is the seller of record, so it should be a named insured, not relying on a brand's vendor endorsement.
2. Product recall or contamination pays the cost of withdrawing, destroying and replacing affected stock, and sometimes lost gross profit. It is underwritten separately and usually needs a defined trigger such as a regulator-ordered withdrawal or confirmed contamination.
3. Non-damage business interruption pays lost gross profit when a site closes without physical damage. A public-authority closure extension is the relevant form, but many wordings limit it to closure for disease or contamination and exclude closure caused by the insured's own breach of regulations.

Each suspended store in Hyderabad is a small loss on its own. The exposure grows when a state agency inspects many stores in a single drive and suspends several at once, which is what the Hyderabad drive did. Operators should ask how per-location limits and aggregates would respond to a multi-store suspension in one city.

The inventory shift does not only affect platforms. It changes the position of the brands supplying dark stores and the landlords leasing space to them.

D2C brands

When a platform buys stock outright, the brand's goods are no longer on consignment at the dark store, and the brand's stock policy will usually stop at the point of sale to the platform. Brands should check where title passes in the supply agreement and align transit cover to it. The manufacturer stays exposed to product liability under the Consumer Protection Act, 2019, which also reaches product sellers, and platforms will increasingly ask for vendor endorsements, higher limits and recall cover as a supply condition. Our guide on D2C brand product liability covers the base programme.

Warehouse and dark-store landlords

A landlord's building policy does not cover the tenant's stock, and the tenant's licence suspension is not the landlord's insured loss. The practical risks for landlords are a vacant or under-used unit if an operator closes a node after repeated suspensions, and higher fire load or cold-room usage than the building was underwritten for. Landlords should confirm occupancy declarations with their insurer and ask tenants for evidence of their own stock and liability cover.

For a platform moving to an inventory model, or already there, the programme review should cover the following before the next renewal:

1. Quantify shrinkage by cause. Split losses into expiry, in-store damage, transit loss, external theft and internal theft. Only the last three and large damage events are insurance questions.
2. Move stock onto a throughput basis. Replace per-location stock limits with stock-throughput or a declaration policy that follows title from brand dispatch to delivery.
3. Fix the rider leg. Confirm in writing whether goods in rider custody are covered, and under which policy.
4. Rewrite the crime cover. Bring gig and third-party workers into the employee definition and negotiate the inventory-shortage exclusion where scan-level evidence exists.
5. Name the platform on product liability and recall. As seller of record, the platform needs its own cover, with limits sized for a multi-store event in one city.
6. Read the BI closure wording. Check whether a regulator's licence suspension triggers non-damage BI, and whether the breach-of-regulation exclusion defeats it.
7. Keep expiry out of the insurance budget. Treat it as a forecasting and markdown problem, not a claim.

The Hyderabad raids and Blinkit's own disclosure point in the same direction. Owning the stock makes the platform responsible for its condition, its location and its safety, and insurance covers only part of that.