Multi-Warehouse Development for E-Commerce on Laravel

CREATE TABLE warehouses ( id BIGSERIAL PRIMARY KEY, name VARCHAR(255) NOT NULL, code VARCHAR(50) NOT NULL UNIQUE, address TEXT, is_active BOOLEAN NOT NULL DEFAULT true, priority INTEGER NOT NULL DEFAULT 0, type VARCHAR(50) NOT NULL DEFAULT 'internal' ); CREATE TABLE warehouse_stock ( id BIGSERIAL PRIMARY KEY, warehouse_id BIGINT NOT NULL REFERENCES warehouses(id), variant_id BIGINT NOT NULL REFERENCES product_variants(id), stock_qty INTEGER NOT NULL DEFAULT 0, reserved_qty INTEGER NOT NULL DEFAULT 0, UNIQUE (warehouse_id, variant_id), CHECK (stock_qty >= 0), CHECK (reserved_qty >= 0) ); CREATE MATERIALIZED VIEW product_total_stock AS SELECT variant_id, SUM(stock_qty) AS total_stock, SUM(reserved_qty) AS total_reserved, SUM(stock_qty - reserved_qty) AS available_qty FROM warehouse_stock GROUP BY variant_id; 

The materialized view is refreshed after each change via trigger or REFRESH MATERIALIZED VIEW CONCURRENTLY scheduled every minute. Our experience shows this approach is 3x faster than live queries on every order—the average query takes 200ms. In 90% of cases, orders are shipped from the nearest warehouse within 24 hours.

How to Choose the Shipping Warehouse?

The core business logic of multi-warehouse is the WarehouseRouter algorithm. We make it configurable:

For most projects, proximity + priority fallback is sufficient. If you're unsure which strategy to choose, contact us—we'll help select the optimal one.

Example router implementation:

class WarehouseRouter { public function resolve(OrderItem $item, Address $destination): Warehouse { $candidates = WarehouseStock::query() ->where('variant_id', $item->variant_id) ->whereRaw('stock_qty - reserved_qty >= ?', [$item->qty]) ->with('warehouse') ->get() ->filter(fn($ws) => $ws->warehouse->is_active) ->sortBy([ fn($a, $b) => $this->byProximity($a->warehouse, $destination) <=> $this->byProximity($b->warehouse, $destination), fn($a, $b) => $a->warehouse->priority <=> $b->warehouse->priority, ]); return $candidates->first()?->warehouse ?? throw new NoWarehouseAvailableException($item->variant_id); } } 

Shipment Management and Reservation

If items from one order are available at different warehouses, the order is split into multiple shipments (Shipment). The customer sees a single order, but internally the system creates distinct tracking numbers. The order status is aggregated: 'completed' only when all shipments are delivered. Order splitting occurs in 15% of orders on average.

Reservation must be atomic—lock a specific warehouse, not just a variant. The reservation operation: UPDATE warehouse_stock SET reserved_qty = reserved_qty + :qty WHERE variant_id = :vid AND warehouse_id = :wid AND (stock_qty - reserved_qty) >= :qty. Simultaneously, insert into stock_reservations. This prevents double-reserving the same item.

How to Synchronize Stock with External Systems?

Each warehouse can have its own stock update channel. For each source, a separate adapter implements the WarehouseStockProvider interface:

• Internal warehouse: WMS via REST API or file exchange (XLSX, CSV).
• Dropshipper: their API with rate limits, often non-standard format.
• Offline store: POS system (1C:Retail, iiko).
• MoySklad: REST API with webhook support.

Synchronization runs independently on a schedule: * * * * * php artisan stock:sync --warehouse=central, etc. Each adapter has its own timeout and error handling.

BOPIS: Pickup from Store

If BOPIS is enabled, the customer selects a pickup point. An API returns available_qty for each item at each warehouse. The frontend builds a list of points and a map (Leaflet, Yandex.Maps), filtering only those with the entire order available. BOPIS implementation typically costs $1,500–$2,000 extra and improves customer satisfaction by 25%.

Warehouse Reporting

Analytics that operations managers actually need:

• Stock per warehouse by category.
• Turnover (sold_qty / avg_stock over period).
• Transfers between warehouses (transfer orders).
• Deficit forecast based on daily sales.

Example turnover query:

SELECT w.name AS warehouse, pv.sku, ws.stock_qty, COALESCE(sales.sold_30d, 0) AS sold_30d, CASE WHEN COALESCE(sales.sold_30d, 0) = 0 THEN NULL ELSE ROUND(ws.stock_qty / (sales.sold_30d / 30.0)) END AS days_of_stock FROM warehouse_stock ws JOIN warehouses w ON ws.warehouse_id = w.id JOIN product_variants pv ON ws.variant_id = pv.id LEFT JOIN ( SELECT variant_id, SUM(qty) AS sold_30d FROM order_items oi JOIN orders o ON oi.order_id = o.id WHERE o.completed_at >= NOW() - INTERVAL '30 days' GROUP BY variant_id ) sales ON sales.variant_id = pv.id; 

Implementation Process

We follow a structured implementation process with clear stages:

1. Analysis: Gather requirements: number of warehouses, types, business processes → Technical specification
2. Design: DB schema, routing algorithms, integration adapters → Architectural document
3. Implementation: Code in Laravel, tests, migrations → Code repository
4. Integration: Connect one or two external providers → Working synchronizations
5. Testing: Load testing, module and integration tests → Test report
6. Deployment: Configure queues, schedules, monitoring → Access and documentation
7. Training: Train managers on the system → Video guides and knowledge base

What's Included

After implementation, you receive:

• Architectural document and DB schema description.
• Full source code repository with comments.
• Configured queues and task schedulers.
• API integration documentation.
• Monitoring access (log centralization, alerts).
• Team training (video guides and knowledge base).
• Warranty on implemented functionality.

Timeline and Cost

• Multi-warehouse schema + priority routing: 5–7 days.
• Order splitting + shipment management: 3–5 days.
• Integration with one external source (1C, MoySklad): 3–5 days.
• BOPIS with pickup map: +3–4 days.
• Analytical warehouse reports: +2–3 days.

A full multi-warehouse system for a store with 2–5 storage points: 2–4 weeks. Average cost ranges from $5,000 to $15,000 depending on complexity. For a typical 3-warehouse setup, the cost is around $8,000 and yields monthly shipping savings of $2,500. Cost is calculated individually—to get an accurate estimate for your project, send us your requirements; we will analyze your warehouses and propose a solution. We offer a warranty on all implemented functionality.

Order a turnkey multi-warehouse implementation with us—get a consultation and an accurate estimate for your project. Contact us to get started.