What Distributed Inventory Does to Your Working Capital (2026)

Every minute you take off a delivery promise gets paid for twice. Once in dark store rent and rider time, which shows up on the P&L. And once in inventory, which shows up on the balance sheet, usually after the plan has been approved.

The second bill is now visible in how consumer brands fund growth. Mint reported in January 2025 that brands borrowing from one revenue-based financier mainly used the money to expand inventory for faster dark store replenishment, and that the lender's quick commerce disbursals had tripled in four months. With Bain projecting that quick commerce will contribute 45 to 50% of India's incremental e-retail GMV over the next five years, that pressure is not going away.

This piece looks at quick commerce working capital from the brand's side of the ledger: why spreading stock across dark stores needs more inventory in total, which levers move that number most, what the standard formula misses, and how the cash cycle compares with selling through a platform.

Why does spreading stock across dark stores need more inventory in total?

Because safety stock does not pool across stores. One city warehouse absorbs demand swings from the whole city: a busy evening in one neighborhood offsets a quiet one across town. Split the same demand across nine dark stores and each store has to buffer its own swings, alone.

The classic statement of this is the square-root law, formalized by David Maister in the International Journal of Physical Distribution in 1976. Under its assumptions, total inventory rises with the square root of the number of stocking locations. Move from one location to nine and safety stock roughly triples. It does not rise ninefold, but it does not stay flat either.

Treat the law as a ceiling, not a constant. It assumes demand at each location is independent. When demand moves together (a festival weekend, or a month-end payday that lifts every neighborhood at once), the pooling benefit of a central warehouse shrinks, and so does the penalty for distributing. The law also has documented limits: Bernhard Fleischmann showed in OR Spectrum in 2016 that it does not hold when warehouses are replenished in full truck loads against a fill-rate target.

For dark stores, the ceiling is close to binding. Each store sees small, noisy order counts, which is exactly the independent, high-variance demand the law describes. Plan for the penalty, then work out how much of it you can buy back.

How much extra cash does a dark store network lock up?

More than most finance plans assume, and much more when replenishment is slow. The worked example below shows the shape of the effect for one SKU in one city.

Assumptions (illustrative, not benchmarks): the SKU sells 300 units a day across the city, with a daily standard deviation of 36 units; demand at each dark store is independent; the brand targets a 95% cycle service level (z = 1.65); the city warehouse is replenished weekly with a five-day lead time; stock in transit between the warehouse and stores averages a quarter of a day's demand; landed cost is ₹400 a unit. Safety stock is z × σ × √(review period + lead time), and cycle stock is half of each replenishment.

Illustrative calculation by Zippee. Your demand variability, lead times and service levels will change the absolute numbers; the relative effects are the point.

Two things stand out. First, the dark store layer is additive. The city warehouse does not shrink when stores are added, because it still holds the cycle and safety stock that protect against supplier lead time. Second, these are per-SKU, per-city figures. An extra ₹1.9 lakh on one SKU is small. Multiplied across every SKU you place in stores and every city you serve, it becomes a board-level number well before the network looks large.

Which lever moves working capital most: store count, replenishment cadence or SKU depth?

Replenishment cadence, in most cases. In the example, consolidating from nine stores to four saves about 7 percentage points of extra inventory. Letting replenishment slip from daily to every three days costs about 36 points. The first decision gets debated in planning meetings. The second is usually made by default, inside a fulfilment partner's operating model, and rarely reviewed by finance.

The arithmetic explains why. Cycle stock grows in direct proportion to the time between replenishments, and safety stock grows with the square root of review period plus lead time. A slower cadence inflates both, in every store, for every SKU. A faster one shrinks both at once.

Store count is less of a free choice than it looks, because the delivery promise sets it. A 60-minute promise needs stock within a short radius of demand; same-day can run from fewer, larger nodes. That makes every speed tier a working capital decision, and it is far cheaper to make SKU by SKU than brand-wide.

SKU depth is where the two meet. Fast movers can sit in every store, because their demand is large enough that the buffer is small relative to sales. The long tail is better served same-day from a city node, where pooling does the work. The tiering logic is covered in Zippee's inventory positioning guide; the point here is that it is also a cash decision. Sharper store-level forecasts shrink σ, the other input in the formula, which is the subject of demand forecasting at the dark store level.

What does the safety stock formula miss?

Stock that is on hand but not sellable where the demand is. The formula assumes every unit in a store is available to the next customer. In a real network, part of it is stranded.

1. Wrong-store stock. Units placed for demand that did not show up. A customer 40 minutes away cannot buy them, so they protect nobody's service level.
2. Short-dated stock. In categories with shelf life, slow sell-through at one store turns into a write-off rather than a transfer.
3. Returns waiting at the store. Units that come back into a dark store and sit there until they are inspected and moved.

The useful metric is available-to-promise against on-hand, tracked store by store rather than network-wide. A network inventory figure can look healthy while several stores hold stock nobody nearby wants and others run dry. On quick commerce platforms, that second failure costs more than the lost order, as covered in why a product vanishes from the shelf after one stockout.

Budget a stranded-stock allowance alongside safety stock, and ask any fulfilment partner how transfers between stores and pullbacks to the hub are triggered, how fast they happen and what they cost.

Is selling through a platform easier on cash than running your own distributed inventory?

Sometimes, on paper. The comparison that matters is the cash conversion cycle (days of inventory, plus days of receivables, minus days of payables), not inventory days alone.

At an Inc42 summit panel in August 2026, one investor argued that quick commerce can pay brands within roughly 15 days, against receivables of 150 to 180 days in traditional and modern trade, which can push scaled brands towards a negative working capital cycle. That holds when purchase orders flow steadily. The platform decides how much to buy, when and for which stores, so the stock you fund upstream is sized to its purchase cadence rather than your demand signal. Listing also turns your packaging and delivery SOPs into suggestions rather than standards.

On your own site, prepaid orders bring cash in at the time of order, and the distributed inventory is yours to place, move and rebalance. Cash-on-delivery orders add a remittance gap that depends on your fulfilment contract.

Most brands will run both. The working capital question is which orders belong where, covered in which orders to fulfil yourself and which to leave to the platform, and what the own-channel route costs to run, covered in what same-day delivery costs on your own site.

How Zippee runs distributed inventory for brands

Zippee runs dark store infrastructure for consumer brands that sell on their own websites, with a network of 140+ dark stores across 21+ cities. It is the operating model behind Quick Commerce-as-a-Service: the brand keeps its storefront and its standards, and Zippee runs the stores, the pick-and-pack and the delivery.

For a finance team, three operating facts matter more than the network map: how often each store is replenished from the city hub, how quickly stranded stock is rebalanced or pulled back, and how fast cash from delivered COD orders is remitted. In Zippee's network: Inventory replenishment is aligned with SKU-level demand and store-level stock requirements. Inter-store transfers and inventory pullbacks help rebalance stock across locations, address demand variations and minimise excess inventory.

Placement starts from demand data rather than a static SKU list. When Clinikally moved to Zippee, SKU-level demand analysis was used to set the inventory held across Zippee's dark stores, and the brand recorded an 81% reduction in RTO rates for the pin codes and SKUs Zippee serviced. Fewer failed deliveries also means less stock tied up in reverse transit, which is a working capital result as much as a logistics one.

Effective inventory planning combines SKU-level demand forecasting with store-level stock visibility to maintain product availability while minimising excess inventory and working capital lock-in.

The bottom line

Distributed inventory is not a cost to avoid. It is the price of a delivery promise, and it can be paid well or badly. Price each speed tier in working capital, SKU by SKU. Treat replenishment cadence as a finance term, not an ops detail. Budget for stranded stock. Compare channels on the full cash cycle. Done this way, the cash your network locks up becomes a decision you made, not a number you inherited.

If you're ready to turn your fulfilment into a competitive advantage, join our waitlist.