Multi-Warehouse Inventory Management: A Practical Guide for D2C & FMCG Brands

Why Multi-Warehouse Inventory Gets Hard

With one warehouse, a spreadsheet almost works. The ops lead knows roughly what is on the floor, the picker knows where things are kept, and when a count looks wrong someone walks over and checks. With three warehouses — say an own facility in Haryana plus 3PL nodes in Bhiwandi and Bangalore — none of that transfers. The same SKU can be ageing toward expiry in one warehouse, selling out in another, and sitting inside a truck between the two. Nobody can walk over and check.

The failure pattern is predictable. Each warehouse keeps its own sheet, updated at its own rhythm. Stock levels are reconciled over WhatsApp. A Blinkit PO gets approved against stock that was actually dispatched yesterday, the shortfall surfaces at picking, the line gets cut, and the cut shows up weeks later as a fill-rate penalty. Meanwhile a batch of the same SKU crosses 50% of its shelf life in the other warehouse because nobody was watching its expiry date at that location.

Step 1: Model the Network — Warehouse Locations, Rack Locations, Case Sizes

Everything downstream depends on master data. In FilFlo, each physical facility — your own warehouse or a 3PL ship-from node — is a Warehouse Location. Inside each warehouse, storage is broken into rack locations. A rack is not decoration: it is the address on every ledger entry and every picklist line. When the system later says "pick 12 cases of the 500ml SKU from rack A-14, batch expiring November," that instruction only exists because the rack was modelled up front.

The product master carries the fields that make warehouse math work: SKU code, EAN, HSN and tax rate, MRP, unit of measure, and — critically for B2B dispatch — case size. FilFlo tracks B2B quantities in case units and converts to unit items (quantity × case size) where the channel needs units, such as GRN entry. Get case size wrong on the master and every downstream count is wrong by a multiple.

Step 2: Batch, MFG/EXP, and Shelf-Life Buckets per Warehouse

For an FMCG brand, stock without a batch identity is a liability. Every inward into a FilFlo warehouse — whether entered manually, received against a procurement PO, or bulk-loaded by CSV — records the batch, manufacturing date, and expiry date alongside the rack. The Inventory Inward log keeps the full history: what came in, into which rack, against which reference, entered by whom.

Batch tracking earns its keep through shelf-life buckets. FilFlo compares each batch's MFG/EXP dates against the SKU's total shelf life and classifies the remaining life into four bands — 0–25%, 25–50%, 50–75%, 75–100% — per SKU, per warehouse. That table answers the question a supply planner actually asks: which stock, in which city, needs to move before it ages out? Quick-commerce buyers enforce acceptance windows on remaining shelf life, so a batch sliding toward the 25–50% band is a candidate to push to Blinkit or Zepto now, at full price, rather than write off later. Batches close to expiry show their EXP dates flagged in red, and a single order line can split across batches with different MFG/EXP dates when that is what it takes to fill it.

This is the working definition of FEFO/FIFO inventory software for food and beverage: first-in-first-out allocation by default (covered next), with expiry visibility per warehouse so the ageing exceptions get handled deliberately instead of being discovered at a channel's receiving dock.

Step 3: FIFO Picking and an Inventory Ledger That Cannot Drift

When an approved B2B order becomes a picklist, FilFlo generates FIFO rack allocations by inward date: the oldest inwarded stock is reserved first, across whatever racks and batches it sits in. The picker gets rack-by-rack instructions; if there is not enough stock, the shortfall is recorded per line — requested, allocated, shortfall — instead of being silently absorbed. Scan-verified picking modes (scan the EAN per line, or scan-first) exist for warehouses that want the extra guardrail.

Underneath all of it sits the inventory ledger, which enforces one invariant at every warehouse: total = available + allocated, tracked at product + rack + batch grain. Allocation moves stock from available to allocated; dispatch writes the outward; a cancelled picklist deallocates. Every one of those movements — inward, allocation, outward, deallocation, adjustment — lands in the Inventory Logger with the available-after balance and the person or process that caused it. When a number looks wrong, you do not re-count first; you read the ledger.

Step 4: QR-Coded Cases and Dispatch Scan Verdicts

Picking errors are cheap to make and expensive to discover — a wrong case that leaves the warehouse comes back as a GRN mismatch, a deduction, or an RTO. FilFlo closes that gap at the dock. Physical cases carry QR codes, and each case moves through its own status chain: available → allocated → picked → dispatched.

At dispatch, the loader scans each case with a camera scanner (or keys in the case code manually). Every scan returns an immediate verdict:

A successful dispatch scan flips the case to dispatched and writes the outward ledger entry in the same motion. Across multiple warehouses this matters doubly: the wrong-case error you prevent in Bhiwandi is one your Bangalore team never has to explain on a GRN call.

Step 5: Cycle Counts with Maker-Checker Approval

Even a disciplined warehouse drifts from its system numbers — damage, pilferage, miscounts at inward, expiry write-offs. The fix is not an annual stock-take that shuts the warehouse for two days; it is a rolling cycle count, and in a multi-warehouse setup it is the main instrument HQ has for trusting numbers it cannot physically see.

FilFlo's Count Adjustment flow is deliberately two-person. The counter enters the physical quantity for a product at a rack; the system computes the variance against the ledger automatically and requires a typed reason — damage, theft, count_error, expiry, other. The adjustment is submitted as pending; a second, senior user approves or rejects it, with before/after snapshots preserved. That maker-checker gate is what stops "adjust the number until it matches" from becoming warehouse culture. Over a quarter, the variance-reason distribution per warehouse is itself a management report: a site whose adjustments skew to theft or damage has a different problem from one that skews to count_error.

Step 6: Stock Transfers Between Warehouses — Gatepasses and In-Transit Ageing

Sooner or later the DOH numbers will tell you to move stock: 68 days of cover in Delhi, 6 in Bangalore. In FilFlo, non-order stock movements — transfers, samples — leave a warehouse through a gatepass, with its own lifecycle: Pending → Approved → Allocated → Picked. Nothing exits the building on a verbal instruction; the gatepass is the authorization, and the outward hits the source warehouse's ledger the moment it is picked.

While the truck is on the road, the shipment appears in In-Transit Ageing — the report that stops in-between stock from disappearing from everyone's mental model. In-transit visibility also protects your buying: FilFlo's replenishment math counts on-hand plus in-transit before recommending a purchase, so you do not double-buy stock that is already moving toward the warehouse that needs it. On arrival, the receiving warehouse inwards the goods with batch and rack details, and the network picture is whole again.

Step 7: 3PL Inventory Visibility Without Rip-and-Replace

Most growing brands do not own all their warehouses — they rent capacity from 3PLs, and each 3PL runs its own WMS. The practical question is not "which single system does everyone use" but "how does the brand keep one operating picture across systems it does not control." FilFlo handles this with two patterns, both live in production deployments:

There is a third, lower-tech pattern that matters just as much: people. 3PL staff can work inside the brand's FilFlo workspace with scoped roles — they see and act on their warehouse only — and a scoped API key lets a 3PL confirm delivery or enter GRN programmatically without getting access to anything else. The brand's ops team and the 3PL's floor team end up looking at the same order, the same picklist, the same ledger.

Step 8: Read the Warehouse Insight Report — Days On Hand per Warehouse

With the mechanics in place, the network becomes readable. The Warehouse Insight report lays out, per warehouse: stock on hand, quantities billed, quantities ordered, and DOH — Days On Hand. DOH is the number that makes warehouses comparable: 300 units might be three weeks of cover in Delhi and three days in Mumbai, and only DOH tells you which.

The same consumption math drives buying. FilFlo's Procurement Alerts compute days left per SKU from actual run-rate — an alert reads "10.5 days left" with a suggested, editable order quantity and the default supplier's last price. The safety floor behind it is deliberately simple and inspectable: 2 × inward TAT × daily run-rate. If your supplier takes 15 days to deliver and the SKU sells 40 units a day, the floor is 1,200 units; when on-hand plus in-transit falls below two inward-TATs of demand, it is time to reorder, rounded up to MOQ or pack size. No black box — a formula your ops head can check on a napkin.

Read together, the reports settle the classic multi-warehouse arguments with data: whether to transfer or reorder (compare DOH against inward TAT — transfers fix imbalances faster than new POs), which warehouse gets constrained stock (the one whose DOH runs out first against committed orders), and which ageing batch goes to quick commerce this week (the shelf-life bucket table, per warehouse).

Putting It Together: A Typical Multi-Warehouse FilFlo Setup

Here is how D2C and FMCG brands with multiple warehouses typically structure their FilFlo workspace for B2B distribution:

Role scoping means each warehouse team only sees and acts on its own location's orders and stock — no risk of one site approving or dispatching another site's orders — while HQ keeps the consolidated view for reporting, replenishment, and transfer decisions. Brands like Jimmy's Cocktails run this pattern across state warehouses feeding modern trade and quick commerce; Anveshan runs it from its own factory through multiple 3PL fulfilment nodes.

A Working Checklist for Multi-Warehouse Discipline

• ✓Every inward carries batch, MFG date, expiry date, and a rack — no anonymous stock.
• ✓Picklists allocate FIFO by inward date; shortfalls are recorded per line, never absorbed silently.
• ✓Dispatches are scan-verified at case level; a wrong_picklist verdict at the dock beats a GRN mismatch two weeks later.
• ✓Cycle counts run weekly per zone with maker-checker approval and typed variance reasons.
• ✓Inter-warehouse transfers move on gatepasses and stay visible in In-Transit Ageing until inwarded.
• ✓Replenishment and transfer decisions start from DOH per warehouse and the 2 × inward TAT × run-rate safety floor.

Frequently Asked Questions