NETNetwork designStop 2 of 11
- What network design decides
- Six distribution network designs
- Facility location models
- Greenfield studies
- Manufacturing footprint & plant roles
- Channel design & omnichannel
- Cross-docking vs stocking
- Postponement in the network
- Reverse logistics networks
- Carbon-aware network design
- Scenario modeling & digital twins
Six distribution network designs
From drop-shipping to store pickup, the six ways to store and deliver products, and which one fits which kind of product.
Read atChange any time
Every distribution network answers two questions: where is the stock held (manufacturer, distributor or retailer) and how does it reach the customer (shipped, delivered to the door, or collected)? Combining the answers gives six classic designs:
| Design | How it works | Fits |
|---|---|---|
| 1. Manufacturer storage, direct shipping (drop-ship) | Retailer takes the order; the manufacturer ships straight to the customer | High-value, low-volume, many variants: a rare camera lens |
| 2. Drop-ship with in-transit merge | Parts from several sources are combined en route into one delivery | Multi-part orders: a PC with monitor and keyboard |
| 3. Distributor storage, parcel delivery | Regional warehouses hold stock and ship by courier | Medium movers that need faster delivery than drop-ship |
| 4. Distributor storage, last-mile delivery | The distributor’s own vans deliver to the door | Bulky or frequent items: groceries, furniture |
| 5. Manufacturer or distributor storage, customer pickup | Orders go to pickup points or lockers | Online orders where customers accept collecting |
| 6. Retail storage, customer pickup | Stock sits in stores; customers buy or collect there | Fast-moving, everyday items: milk, shampoo |
The pattern behind the table: the further upstream you hold stock (designs 1 and 2), the more you pool inventory, but the slower and costlier each delivery is. The closer to the customer (design 6), the faster and simpler the hand-over, but the more inventory sits in many places.
So match the design to the product. A ₹2 lakh camera that sells a few units a month belongs upstream, with one shared stock. Milk belongs in the corner store. Most companies run several designs at once, one per product segment.
Words to use
- Drop shipping
- The manufacturer ships directly to the end customer; the retailer holds no stock.
- In-transit merge
- Combining items from different sources into one delivery on the way to the customer.
- Last-mile delivery
- The final leg to the customer’s door; usually the costliest per unit.
How the designs compare
| Design | Inventory | Transport cost | Facility and handling | Response time | Product variety |
|---|---|---|---|---|---|
| 1. Drop-ship | Lowest (fully pooled) | High (single parcels over long distances) | Low | Slow | Very high |
| 2. Drop-ship with in-transit merge | Lowest | Somewhat lower than 1 | Low, but coordination is hard | Slow | Very high |
| 3. Distributor storage, parcel | Low to medium | Lower than drop-ship | Medium | Medium | High |
| 4. Distributor storage, last mile | Medium | Highest (door delivery) | Medium to high | Fast | Medium |
| 5. Customer pickup points | Low to medium | Low (consolidated drops) | Medium (pickup sites) | Medium | High |
| 6. Retail storage | Highest (stock everywhere) | Lowest per unit | High (stores) | Immediate | Lower |
The pattern is the same one as risk pooling: the further upstream the stock, the less of it you need, and the more each delivery costs.
A ₹2 lakh lens sells about 2 units a month across 50 stores.
- In every store (design 6): each store needs at least 1 unit to be able to sell it: 50 units, ₹1 crore of stock.
- Held centrally (design 1 or 3): about 4 units cover the lead time and a buffer: ₹8 lakh of stock.
- The 46 units saved are worth ₹92 lakh. At a 25% holding rate that is ₹23 lakh a year, against roughly 24 parcel deliveries a year to customers, a few thousand rupees each at most.
For milk the arithmetic flips: it sells every hour, a customer won’t wait a day, and a parcel per litre would cost more than the milk.
Segmenting by product and customer
Most companies run several designs:
- Fast movers: stocked close to customers (designs 4 and 6).
- Medium movers: regional warehouses shipping parcels (design 3).
- Slow, expensive or highly customized items: drop-shipped or held centrally (designs 1 and 2).
- Online orders: home delivery or pickup points (designs 3, 4 and 5).
Information makes upstream designs work
Drop-shipping and in-transit merge depend on information more than buildings: the retailer must see the manufacturer’s stock in real time to promise orders, and merge-in-transit needs carriers who can combine shipments from several sources into one delivery on time. Without that visibility, upstream designs produce broken promises and split deliveries. Chopra’s (2003) framework, which set out these six designs, stresses that the information infrastructure is part of each design’s cost.
Last-mile density
Door delivery (design 4) is expensive per drop, but its cost falls sharply with density: the more drops per kilometre of route, the cheaper each one becomes. That explains why home delivery works in dense cities long before it works in small towns, and why quick-commerce dark stores, small stockrooms placed inside dense neighbourhoods, can deliver within minutes at a manageable cost. Pickup points and lockers (design 5) are a way to buy density: many customers’ parcels go to one drop.
Indian channel realities
- General trade: most consumer goods still reach small kirana stores through distributors who hold stock and deliver small drops on fixed routes (beats). That is a form of distributor storage with last-mile delivery.
- Cash and credit: distributors and retailers expect credit terms, so the distribution design is also a working-capital design.
- E-commerce: cash on delivery and high return rates in some categories add cost to home-delivery designs.
Returns are part of the design
Upstream designs make returns slower and more expensive: a drop-shipped item may have to travel back to the manufacturer. Store-based designs make returns easy for customers but scatter returned stock across many locations. Categories with high return rates, such as fashion, need the reverse flow designed alongside the forward one.
Interview kit
Practise it out loud
Answer each question yourself before opening the model answer.
Distribution designs come from two choices: where inventory is held, at the manufacturer, distributor or retailer, and how it reaches the customer, by shipping, home delivery or pickup. That gives six designs from drop-shipping to retail storage with pickup. Upstream storage pools inventory and suits high-value, slow-moving items; downstream storage costs more inventory but gives fast, cheap hand-over for high-volume everyday products. Most firms combine designs by product segment.
Which design would you use for an expensive camera that sells a few units a month, and which for milk?
The camera: manufacturer or central storage with direct shipping, because pooling one stock across all customers saves a lot of expensive inventory and customers will wait a day or two. Milk: retail storage with customer pickup, because it’s cheap, fast-moving, perishable and needed immediately.
What is the main cost of drop-shipping?
Transport and response time. Each order ships individually from far upstream, so outbound freight per unit is high and delivery is slower. Order visibility and returns are also harder, because the manufacturer and the retailer need tightly connected systems.
Connected stops
Where this idea feeds in, or gets its inputs from.
Sources and further reading
- Chopra, S. & Meindl, P. Supply Chain Management: Strategy, Planning, and Operation. Pearson. Chapter on designing distribution networks and applications to online sales.
- Chopra, S. (2003). Designing the distribution network in a supply chain. Transportation Research Part E, 39(2), 123–140.